- Société géologique de France Éditeur, Paris
The Lower Carboniferous Baralacha La basaltic dykes were emplaced along transtensional faults. The basalts exhibit tholeiitic and alkaline affinities. The tholeiites are TiO2-poor, moderately enriched in light rare earth (LREE), and display Nb and Ta negative and Th positive anomalies. The alkali basalts, compared to the tholeiites, have higher TiO2, rare earth and highly incompatible trace element contents and greater LREE enrichments. The Nd and Pb isotope compositions of the Baralacha La basalts suggest that they derive from the partial melting of an enriched OIB mantle source, characterized by a HIMU component, and contaminated by the lower continental crust. The Baralacha La dyke swarm represent the remnants of an early rifting event on the northern Indian passive margin.
A major change in the geographical distribution of the lithospheric plates occurred during the late Paleozoic [Nikishin et al., 2002]. This change corresponds to the transitional period between the development of Pangea by accretion of lithospheric fragments, island arcs linked to the closure of oceanic basins and its break-up during Permian-Triassic times [Stampfli et al., 2001; Nikishin et al., 2002]. The Neotethyan opening took place also during Permian-Triassic times, leading to the quick northward drift of the Cimmerian blocks [Sëngor, 1984], detached from the northern margin of the Indian-Arabian plate. The Neotethyan opening began during the late Carboniferous-early Permian, east of Australia and propagated towards the west, along the Indian and Arabian plates during the middle to late Permian [Stampfli et al., 2001].
This geodynamic reorganization of oceanic and continental realms was accompanied by the emplacement of large igneous provinces (LIPS), which is attributed to an important thermal activity and a reorganization of the mantle convecting system [Nikishin et al., 2002]. Among the best documented Permian LIPS, one finds: (i) the Siberian [Arndt et al., 1998], Emeishian-Song Da [Xu et al., 2001] and Panjal Traps [Gupta et al., 1983], (ii) the magmatism of the Hawasina nappes in Oman [Béchennec et al., 1991; Maury et al., 2003; Lapierre et al., 2004]. According to Abbott and Isley , the Paleozoic LIPS developed during the Devonian (377–368 Ma) and Permian (254–252 Ma) times and the duration of their emplacement did not exceed 12 Ma (12 ± 3 Ma). Tectono-magmatic activity began to develop along the Gondwana eastern margin in the Middle Carboniferous [Stampfli et al., 2001] likely leading to the break-up of the Cimmerian blocks. On the northern Indian plate, a Carboniferous mafic volcanism was emplaced along extensional, locally synsedimentary faults. The latter represents evidence of the first extension occurring on the future Indian northern margin. Very few geochemical and especially isotopic data on the Carboniferous magmatism developed on the Gondwana margins are available in the literature. The only study published on the Baralacha La basaltic dykes is based on major and few trace elements [Vannay and Spring, 1993]. That is why we have developed new geochemical (major, trace element, Nd, Sr and Pb isotopes) investigations on the Baralacha La basalts and their clinopyroxenes.
The Himalayan Ranges result from the collision of the Indian and Asian plates [Gansser, 1964]. The initial convergence led to the development of an Andean-type margin during the Cretaceous and lower Paleocene as the result of the subduction of an oceanic realm (the Neotethys) beneath southern Asia (Tibet). In contrast, the Indian northern margin remained passive [Bassoullet et al., 1980; Baud et al., 1982]. After the lower Eocene, the Indian northern margin was shortened in a series of tectonic slices, which are nowadays caught between the Upper Himalayan crystalline domain to the south and the Indus-Zangbo suture to the north [Frank et al., 1977]. These dismembered units of the Indian northern passive margin are largely exposed in the Zanskar, Spiti and Upper Lahul areas of northwestern Himalaya (fig. 1⇓) [Srikantia and Barghava, 1982; Baud et al., 1982; Fuchs, 1982].
The “pre-rift units” of the Indian northern margin consist of platform-type detritic sediments, the age of which ranges from the late Precambrian up to the Lower to Middle Permian [Bagati, 1990; Spring, 1993; Vannay, 1993]. The “syn-rift” units (Upper Permian to Middle Triassic) are composed of platform-type limestones and flyschs [Lamayuru Unit, Bassoullet et al., 1980; Cannat and Mascle, 1990] deposited on the shelf and slope, respectively. The “post-rift” units (Upper Triassic to Lower Eocene) consist of platform-type limestones (Triassic-Dogger, Upper Cretaceous-Paleocene) or detritic sediments deposited on the shelf (Upper Cretaceous) while pelagic sediments deposited on the slope [Garzanti et al., 1987]. All these sedimentary units have been metamorphosed during the collision [Colchen et al., 1986; Steck et al., 1993; Steck, 2003].
Four igneous events are present within the “pre-rift units” [Spring et al., 1993; Vannay and Spring, 1993, i. e. (i) Cambrian-Ordovician granites [Le Fort et al., 1986; Girard and Bussy, 1999], (ii) Carboniferous Baralacha La basaltic dykes [Vannay, 1993] (figs. 2⇓ and 3⇓), (iii) early Permian alkaline granite and microgabbro intrusions [Spring et al., 1993], (iv) Middle Permian Panjal Traps [Bagati, 1990; Gaetani et al., 1990].
The Baralacha La basaltic dykes were emplaced along transtensional, locally synsedimentary faults, that never affect formations younger than the Lower Carboniferous (fig. 3⇑) [Vannay, 1993]. A few exposures of conformable basaltic flows occur within the evaporites of the lowermost Carboniferous (Lipak Formation). The late Carboniferous is locally absent and the Paleozoic ends with Permian detritic sediments that overlie unconformably the older units.
In order to better constrain the genesis of the Baralacha La basalts, new investigations were undertaken based on the trace element of eleven rocks (table II⇓⇓) sampled by one of us (Vannay). Their location is shown in figure 2⇑. Clinopyroxenes of four samples (V148, V300, V326 and B50) were analyzed for major and trace element chemistry (table I⇓). Nine among the best preserved samples were analysed for isotope compositions (Nd, Sr, Pb, table II⇓⇓).
Before Spring and Vannay , the Baralacha La basalts were considered as the feeder dykes of the Panjal Traps [Fuchs, 1982; Gaetani et al., 1990], which are related to the Permian-Triassic Neotethyan opening [Honneger et al., 1982; Stampfli et al., 1991]. These authors have clearly shown on the basis of stratigraphic and some geochemical data, that the Baralacha La basalts represent remnants of an igneous activity completely different from that of the Panjal Traps.
PETROGRAPHY AND MINERALOGY OF THE BARALACHA LA BASALTIC DYKES
Textures of the dykes
Among the Carboniferous dykes, two textures have been observed: aphyric basalt with intergranular texture and porphyritic basalt.
The intergranular basalts (V300, V326, V324, fig. 2⇑, tabl. I and II) consist of millimeter sized-plagioclase laths replaced by albite, subhedral clinopyroxene and interstitial aggregates of chlorite, actinolite and epidote. Some samples (V324) include also millimeter-sized olivine pseudomorphs, now replaced by chlorite.
The porphyritic basalts differ from the aphyric ones by the presence of centimeter-sized clinopyroxene (B50, V148) or plagioclase (V164) phenocrysts. The intersertal groundmass is formed of tiny plagioclase microlites, now replaced by albite, anhedral oxides, chlorite, albite, epidote and calcite. Locally, clinopyroxene microphenocrysts may occur in the groundmass (B50). Quartz – (V148) or chlorite + pistachite-filled vesicles may occur.
Major elements compositions of clinopyroxenes were determined using a Cameca SX-100 electron microprobe fitted with five spectrometers at the « Service commun Microsonde » (Université de Montpellier). The standard procedures are 20 kV and 10 nA with an electron beam of 1 μm width and integrated counting times of 20–30 s. Synthetic and natural minerals were used as standards. Alkalis were determined first to minimize Na loss during measurements. A computer correction program was used to calculate the element concentrations. The accuracy of major element contents is better than ± 5% of the total values.
Clinopyroxenes of the two aphyric basalts (V300, V326) are light-pink colored and plot in the diopside field or at the boundary between the diopside and augite fields (fig. 4a⇓) (Wo40-47, En33-36, Fs19-22). In contrast, colorless clinopyroxenes of the porphyritic basalts (B50, V148) show augite composition (Wo32-43, En44-49, Fs11-20) (fig. 4a⇓). In the Ti versus Ca + Na diagram [Leterrier et al., 1982, fig.4b⇓], the Ti-rich diopsides (3.4 < TiO2 < 1.6) of the aphyric basalts (V300, V326) cluster in the alkali field, while augites of the porphyritic basalts, characterized by low Ti contents (TiO2 = 0.6) for a wide range of Ca + Na levels plot in the tholeiitic field.
Figure 4c⇑ shows the behavior of clinopyroxene compatible elements within the same sample and/or within a single grain. In the aphyric basalts (V300, V326), clinopyroxene is Cr-poor and its MgO and CaO contents remain rather constant while TiO2 and FeO increase significantly from core to rim and within the same sample. In the porphyritic ones (B50, V148), clinopyroxenes are less FeO-rich but are characterized by a significant FeO and TiO2 enrichment and CaO and Cr2O3 depletion within the same sample. Such an evolution is likely to reflect a differentiation process. Clinopyroxenes from sample B50 are relatively primitive and enriched in Cr2O3, CaO and MgO. The very low Cr2O3 contents of clinopyroxene in V148 indicates the differentiated nature of the lava.
Trace-elements contents determinations on minerals were obtained by laser-ablation ICP-MS mass spectrometry using a 193 nm an Ar-F 193 nm Lambda Physics© Excimer laser coupled with a Perkin-Elmer 6100DRC ICPMS at the Institut of Mineralogy and Geochemistry of the University of Lausanne. NIST610 and 612 glasses were used as external standards, Ca and Si as internal standards after microprobe measurements on the pit sites. Ablation pit size varied from 40 to 60 mm. BCR2 basaltic glass was regularly used as a monitor to check for reproductibility and accuracy of the system. Results were always within ± 10% of the certified values.
The C1 chondrite-normalized [Sun and McDonough, 1989, fig. 5⇓] rare earth elements (REE) patterns of the Baralacha La clinopyroxenes are light (L)REE depleted with respect to the heavy (H)REE [(La/Yb)N ranging between 0.3 and 0.7]. The tholeiitic clinopyroxenes differ from the alkalic ones by their lower REE contents (less than 10 times the chondritic abundance), the lack of HREE depletion with respect to the middle (M)REE, and more or less marked Eu negative anomalies (Eu/Eu* = 0.9).
The primitive mantle-normalized [Sun and McDonough, 1989, fig. 5⇑] multi-element plots of the Baralacha La clinopyroxenes show common Nb, Ta, Zr and Ti negative anomalies. The behavior of Rb, Ba and U is not consistent with the clinopyroxene affinity and can be attributed to the mobility of these lithophile elements (large ion lithophile elements, LILE) during alteration or weathering processes. The Rb mobility is shown by sample B50 clinopyroxene. Indeed, as this rock sample has the most primitive composition, clinopyroxenes in it should have the lowest Rb content. On the contrary clinopyroxenes in B50 have the highest Rb content. Therefore, this is consistent with a remobilization of Rb after crystallization. The alkali clinopyroxenes show a more or less marked Pb negative anomaly while the tholeiitic clinopyroxenes have a small positive Pb anomaly. The Sr positive or negative anomalies could be related to alteration processes. However, the presence of both Pb and Sr negative anomalies in the tholeiitic clinopyroxenes, could reflect early plagioclase removal.
In summary, clinopyroxenes from the Baralacha La show tholeiitic and alkaline affinities and Nb and Ta negative anomalies with respect to La and Th, respectively.
GEOCHEMISTRY OF THE BARALACHA LA BASALTIC DYKES
Major, compatible and incompatible trace elements were determined (table II⇑⇑) by ICP-optical spectroscopy at the Université de Bretagne occidentale in Brest, using the procedure of Cotten et al. . Trace elements including U, Pb and the whole set of REE (table II⇑⇑) were analyzed by ICP-MS at the Université Joseph Fourier in Grenoble, after acid dissolution of 100 mg sample, using the procedure of Barrat et al. . Limits of detection for REE and Y = 0.03 ppm, U, Pb and Th = 0.5 ppm, Hf and Nb = 0.1 ppm, Ta = 0.03 ppm and Zr = 0.04 ppm. Standards used for the analyses were JB2, WSE, BIR-1 and JR1. Analytical errors are 1–3% for major elements and less than 3% for trace elements. All the samples were pulverised in an agate mill. Ta contents in some samples (V192) were not accurately measured.
Sr (static acquisition) and Nd (dynamic acquisition) isotope ratios (tabl. II) were determined on 9 samples at the Laboratoire de Géochimie Isotopique de l’Université Paul Sabatier de Toulouse on a Finnigan MAT261 multicollector mass spectrometer using the analytical procedure of Lapierre et al. . Results on standards yielded 143Nd/144Nd = 0.511850 ± 0.000017 (2σ external reproductibility) on 12 standards analysed. Results on NBS 987 Sr standard yielded 87Sr/86Sr = 0.710250 ± major 0.000030 (2σ external reproductibility) on 11 standard determinations. 87Sr/86Sr and 143Nd/144Nd were normalized for mass fractionation relative to 86Sr/88Sr = 0.1194 and 146Nd/144Nd = 0.7219 respectively. εNdi were calculated with the present day (143Nd/144Nd)CHUR = 0.512638 and (147Sm/144Nd)CHUR = 0.1967. Prior to Nd and Sr separation and acid digestion, two leaching steps in 2.5N HCL during 20 minutes at 100oC were made.
For lead separation, 9 powdered samples were weighted to obtain approximatively 100 to 200 ng of lead. A leaching step with 6N HCl during 30 minutes at 65oC was made before acid digestion. Samples were then dissolved during 36–48 hours on a hotplate with a mixing of tridistilled Hf: HNO3 concentrated acids. After evaporation to dryness, 1 ml of HNO3 was added to the residue and kept at about 90oC for 12–24 h. After complete evaporation, 0.5 ml of 8 N HBr was added to the sample which was kept at 70oC for 2–3 h before complete evaporation. The chemical separation of lead was done using 50 μl of anion exchange resin (AG1X8, 200–400 mesh) and samples were loaded and washed in 0.5 N HBr. Lead was then eluted in 6 N HCl. Pb blanks were less than 40 pg and are negligible for the present analyses.
Lead isotope analyses were made on a VG model Plasma 54 magnetic sector-inductively coupled plasma-mass spectrometer (MC-ICP-MS) at the Ecole Normale Supérieure de Lyon. Lead isotope compositions were measured using the Tl normalization method described by White et al. . For Pb isotope analysis, samples were bracketed between NBS 981 standards and calculated with respect to the value reported for this standard by Todt et al. . This technique yields internal precision of ca. 50 ppm (2σ and an external reproductiblity of ca. 150 ppm [2σ Blichert-Toft et al., 2003] for 206Pb/204Pb ratios determined on 20 NBS 981 standards.
Major element geochemistry
The loss of ignition (LOI) of the Baralacha La dykes ranges between 3 and 6.5% (tabl. II). Considering the alteration, the low grade metamorphism and weathering processes that have affected the Carboniferous dykes, which prevent us to use mobile elements (K2O, Na2O, SiO2, CaO) to characterize the magmatic affinities, we have used the TiO2 contents and the REE patterns of the whole rocks to distinguish the alkali and tholeiitic basalts. The tholeiites (V192, V148, B29, B50) are characterized by low TiO2 contents (LTi, TiO2 < 2%) and slightly LREE enriched patterns while the alkali basalts (V164, V300, V324, V326, V327, S7) have higher TiO2 contents (HTi, TiO2 > 2%), and greater enrichments in LREE. Moreover, the tholeiites have higher MgO and lower P2O5 contents than the alkali basalts.
Trace element geochemistry
The chondrite-normalized [Sun and McDonough, 1989, fig. 6b⇓] REE patterns of the Baralacha La tholeiites are slightly enriched in LREE [(La/Yb)N < 2.9] and show Eu positive or negative anomalies. The alkali basalts differ from the tholeiites by higher REE contents, greater LREE enrichments [3.5 (La/Yb)N < 5.2] and no Eu anomalies.
Because of LILE mobility during alteration and weathering processes, these elements are not shown in the primitive mantle-normalized [Sun and McDonough, 1989, fig. 6⇑] multi-element plots of the Baralacha La basalts. The tholeiites are characterized by Nb and Ta (La/Nb ~ 1.5 with the exception of V148) negative and Th positive anomalies, LREE enrichments and rather flat patterns. These features are those of continental tholeiites. The alkali basalts, compared to the tholeiites, exhibit greater enrichments in incompatible elements, and negative anomalies in U. Some samples (S7, V300) show Zr and Hf negative anomalies; others (V300, V324, V326) weak Nb and Ta negative anomalies (1.06 < La/Nb < 1.13).
The complete isotopic data set has been corrected for in situ decay assuming an age of 350 Ma, based on stratigraphic constraints (refer to the stratigraphic column, fig. 3b⇑). The measured and initial ratios are shown in table II⇑⇑.
All the Baralacha La basalts have high initial 87Sr/86Sr ratios that do not reflect the composition of the sources of the mafic melts. These high 87Sr/86Sr are due to alteration, low grade metamorphism or weathering processes that have affected these rocks.
The εNdi values of the Baralacha La basalts range between + 2.3 and – 1.4. Such values suggest that the composition of the source (or sources) of the Baralacha La basalts resembles that of the Bulk Silica Earth (BSE) and reflects to some extent crustal contamination or assimilation. In order to characterize the importance of this crustal contamination, we have plotted εNdi versus La/Nb and Nb/Th (fig. 7a⇓). Three out of four tholeiites (V148, V192, B29) and three alkali basalts (V300, V326, V33B) out of five show a negative correlation in the εNdi versus La/Nb plot. In the εNdi–Nb/Th correlation diagram, only the tholeiites exhibit a negative correlation; the samples with the lowest (B50) and highest (V148) εNdi values have the highest and lowest Nb/Th ratios, respectively. To determine if the crustal contamination of the Baralacha La basalts is related to assimilation, fractional, crystallization processes (AFC), we have plotted these rocks in an εNdi-Zr (ppm) diagram (fig. 7a⇓), Zr being considered here as fractionation index. In this diagram, solely the tholeiites display a correlation and surprisingly, the rock (B50) with the lowest Zr content (and the highest MgO) has the lowest εNdi value.
In Pb-Pb correlation diagrams (fig. 7b⇓), the Baralacha La basalts, whatever their affinity, plot above the North Hemisphere Reference Line [NHRL, Zindler and Hart, 1986] and are aligned between either BSE and high μ (238U/204Pb ratio, HIMU) fields or BSE and enriched mantle Type I (EMI) fields. Samples B29 (tholeiite) and V300 (alkali), characterized by the highest La/Nb ratios, plot near the EMI field. Inversely, samples V148 (tholeiite) and V333B (alkali) that have the highest εNd values and the lowest La/Nb ratios plot towards the HIMU field. However, the two alkali basalts (V164, V327) that are not located on the εNdi-La/Nb negative correlation trend (fig. 7a⇑) plot at both ends of the trends observed in the Pb-Pb correlation diagrams. Sample V164 with one of the lowest εNdi values (−0.43, tabl. II) plots near EMI. Sample B50 which has the lowest εNd value (−1.4, tabl. II) plots near EMII.
In the εNdi versus (206Pb/204Pb)i diagram (fig. 7c⇑), all the basalts, with the exception of samples B50 and V327, display a positive correlation between HIMU and EMI fields. Located at the two ends of this trend, B29 (tholeiite) and V164 (alkali) are close to the BSE and EMI fields, respectively, while V148 (tholeiite) and V333B (alkali) plot towards the HIMU field. B50 and V327 that differ by higher (206Pb/204Pb)i plot towards the EMII field.
Involvement of the continental crust in the genesis of the Baralacha La basalts
Whatever the affinity of the Baralacha La basalts, their Nb and Ta negative anomalies and Nd and Pb isotopic ratios show that these volcanic rocks derived from the melting of an enriched mantle source (Oceanic Island Basalt, OIB), contaminated to some extent by continental crust.
On the basis of the εNdi-Zr diagram (fig. 7a⇑), we may assume that the variations of the εNdi of the Baralacha La basalts towards 0 and negative values, at least for the tholeiites, are not related to AFC processes.
The correlation shown by the Baralacha La tholeiitic dykes in the εNdi-Nb/Th diagram (fig. 7a⇑) suggests that they have been more or less contaminated by the upper crust because their lowest Nb/Th (B29, B50) ratios approach that of the upper crust [2.38, Weaver and Tarney, 1984]. The Nb/Th ratio of the lower crust [11.90, Taylor and McLennan, 1995] is significantly higher than that of the lower crust. According to the εNdi-Nb/Th diagram, sample B50 appears to be the most contaminated by the upper crust. This is in agreement with its Pb and Nd ratios because this sample plots near the EMII field in the Pb-Pb and εNdi-(206Pb/204Pb)i diagrams. This assumption does not fit with sample V148 because, even if it plots near the EMII and BSE fields in the Pb-Pb diagrams, (i) its Nb/Th ratio is greater than that of the upper crust, and (ii) it plots near the HIMU field in the εNdi and (206Pb/204Pb)i diagram (fig. 7c⇑). Thus, we can assume that samples V148 and B50 derive from the melting of at least two components, an enriched source containing some HIMU component and the upper continental crust (EMII). The isotopic composition of sample V148 could represent that of the enriched mantle source containing some HIMU component. For samples B29 and V192, it seems more difficult to characterize the different sources that contributed to their genesis because, they plot in the middle of the BSE-EMI, BSE-EMII trends in the Pb correlation diagrams. However, the lower crust (EMI) could be involved in the genesis of samples B29 and V192 because they plot towards the EMI field in the εNdi-(206Pb/204Pb)i diagram.
No correlation is observed in the εNdi-Nb/Th diagram (fig. 7a⇑) for the Baralacha La alkali dykes. For more or less similar εNdi values (+ 1.7), the Nb/Th ratio ranges between 5.5 and 6.9. Moreover, sample V164 with the lowest εNdi value (− 0.45) has one of the greatest Nb/Th ratios (8.86). This suggests that the Nb/Th variations are more linked to differences in the degree of mantle partial melting than to the extent of crustal contamination of the mantle source. The upper crust does not seem to be the main crustal contaminant of the alkali dykes because their Nb/Th are higher than that of the upper crust, and closer to that of the lower crust. Moreover, the alkali basalts, in the (208Pb/204Pb)i and εNdi versus (206Pb/204Pb)i diagrams (figs. 7b and c⇑), are located along the BSE-EMI trend (with the exception of V327).
Origin of the geochemical differences between tholeiitic and alkalic basalts
On the basis of correlations between P2O5 (% wt) and Zr (ppm) and the total alkali versus silica diagram, Vannay and Spring  already pointed out the tholeiitic and alkaline affinities of the Baralacha La dykes. They also suggested that the Baralacha La basalts were cogenetic and that crystal fractionation was a major process during their genesis. According to these authors, the most enriched basalts in incompatible elements derived from the most depleted ones by crystal fractionation.
We do not agree that tholeiitic and alkalic basalts are cogenetic and linked by crystal fractionation process. We think, on the basis of elemental and isotopic chemistry, that the differences in incompatible trace element contents between the tholeiitic and alkalic Baralacha La basalts are linked to variations of the mantle partial melting ratios and degrees of enrichment of the sources.
Indeed, no correlation trends are observed in La (ppm) versus (La/Yb)N plot (not presented here) [Caroff et al., 1997]. If the tholeiites and alkali basalts were linked by crystal fractionation, the (La/Yb)N ratio should remain constant while La increases. This is not the case.
Samples V327, V192, and B29 have similar εNdi values (ranging between + 0.2 and 0.45, tabl. II). However, the alkali sample V327 differs from the two others by higher contents in LREE and Th (tabl. II). Similarly, samples V333B and V148 with similar εNdi values (+ 2) differ by incompatible trace element contents. The alkali sample V333B, compared to tholeiitic V148 sample, has a higher (La/Yb)N ratio (3.4) and Th content (2.1 ppm, tabl. II). In both cases, the differences in incompatible trace element contents between tholeiitic and alkaline rocks can be attributed to variations of the mantle partial melting degrees because the sources of both rocks have similar Nd isotopic compositions. Inversely, some tholeiitic basalts (V148, B29, B50, table II⇑⇑) have similar (La/Yb)N ratios (ranging between 2.5 and 2.8) but differ by their εNdi, ranging between + 2.3 and − 1.4 (table. II⇑⇑). In this case, the differences in the incompatible trace element contents of the three tholeiites can be attributed to differences in composition of the sources (mantle versus continental crust) involved in their genesis.
The Baralacha La dykes never intrude rocks younger than early Carboniferous and are associated with contemporaneous transtensional faults, which are locally synsedimentary. These tectonic structures evidence the presence of an early Carboniferous extensional tectonic event in the Tethyan Himalayan Ranges [Vannay and Spring, 1993]. The early Permian granite and alkalic microgabbro of Yunam were emplaced during another likely extensional event of the northern Indian margin. In the Middle Permian, the Panjal continental flood tholeiites were emplaced during a major rifting event that led to the thinning and break-up of the northern Indian margin; the Neotethyan opening and the northward drift of the Cimmerian blocks (Lhassa, Iran).
Thus, three rifting related events associated with extension and magmatism characterize the evolution of the northern Indian margin. The early Carboniferous event predates by about 70 Ma the Permian Neotethyan opening and the emplacement of the widespread volcanism known from the Himalayan Ranges up to Oman and through Iran. Taking into account the important time gap between the emplacement of the Baralacha La basaltic dykes and the Panjal Traps, it seems difficult to consider the Baralacha La basaltic dykes as an early event of the Neotethyan opening. The two tectono-magmatic events (early Carboniferous and early Permian) are most probably related to the thinning and faulting of the northern Indian plate prior to the emplacement of the Panjal Traps during the Middle Permian.
Tholeiitic and alkalic basalts deriving from an enriched mantle source, characterized by an HIMU type component, and contaminated by the lower continental crust (and in some samples by the upper crust) have been emplaced in the early Carboniferous on the northern Indian plate. These basalts are related to a transtensional synsedimentary fault system. This early Carboniferous magmatism was followed some 70 Ma after by the eruption of the Mid to Upper Permian Panjal Traps which are associated to the rifting of the Indian northern passive margin and the Neotethyan opening. The mid to late Permian Panjal Traps can be correlated with the mid-Permian tholeiitic and alkaline volcanics of the Oman Ranges [Maury et al., 2003] considered to represent the igneous remnants of the Arabian margin rifting and Neotethyan opening. This widespread Permian within-plate volcanism is probably related to the melting of a large plume, named the “Tethyan” plume [Lapierre et al., 2004]. Taking into account the important temporal gap between the emplacement of the Baralacha La dykes and the voluminous flows of the Panjal Traps, it is difficult to consider the early Carboniferous tectono-magmatic activity as being related to the Neotethyan opening. However, we assume that the Baralacha La dykes represent: (i) the remnants of a faulting event recorded on the future Indian northern margin and (ii) the upwelling of an enriched asthenospheric mantle during the early Carboniferous but not the initial Permian Tethyan plume.
This work was supported by the Institut National des Sciences de l’Univers, programme « Intérieur de la Terre ». Many thanks to P. Telouk and A. Agranier and the Service commun National du MC-ICPMS of the Ecole Nationale Supérieure de Lyon for their help during the Pb isotope measurements.
- Manuscript Received 14 January 2004.
- Manuscript Accepted 6 April 2005.