Elemental and Nd–Sr–Pb isotope geochemistry of flows anddikes from the Tapi rift, Deccan flood basalt province, India

D. Chandrasekharam a,), J.J. Mahoney b, H.C. Sheth a,1, R.A. Duncan c

a Department of Earth Sciences, Indian Institute of Technology, Powai, Bombay 400 076, Indiab Department of Geology and Geophysics, School of Ocean and Earth Science and Technology, UniÕersity of Hawaii, Honolulu, HI 96822,

USAc College of Ocean and Atmospheric Sciences, Oregon State UniÕersity, CorÕallis, OR 97331, USA

Abstract

The Deccan Traps are a large rift-associated continental flood basalt province in India, parts of which have been studiedextensively in terms of geochemistry, palaeomagnetism and stratigraphy. However, the basalts of the Tapi rift in the centralpart of the province have been little-studied thus far. Two ENE–WSW-trending tectonic inliers of the Deccan basalts in thisregion, forming ridges rising from younger alluvium, are made up of basalt flows profusely intruded by basaltic dikes. Bothof these ridges lie along a single lineament, although they are not physically continuous. The flows are aphyric,plagioclase-phyric and giant-plagioclase basalts, and the dikes are aphyric or plagioclase-phyric. We consider the two inliersto have been originally continuous, from the presence of bouldery remnants of a major dioritic gabbro dike along both.Samples of this dike from both ridges have previously yielded typical Deccan ages of 65.6"0.5 Ma and 65.6"0.6 Ma by

40 39 87 86 Ž .the Ar– Ar incremental heating technique. Initial Srr Sr ratios and ´ t values, and present-day Pb isotopic ratios ofNd

most dikes indicate that they are isotopically similar to lavas of the Mahabaleshwar and Panhala Formations of the WesternGhats, about 450 km to the south. Their mantle-normalized trace element patterns have small Pb and Ba peaks. One dike hasa strong Bushe Formation affinity and a Nd–Sr isotopic composition more extreme than that of any other Deccan rock yet

Ž . Ž87 86 .sampled, with ´ t sy20.2 and Srr Sr s0.72315. Its mantle-normalized element pattern shows large Pb, Th and UNd t

peaks and large Nb–Ta troughs. Its elemental and isotopic chemistry reflects substantial continental contamination. Theflows cut by the Mahabaleshwar-type dikes are isotopically similar to the Poladpur Formation lavas of the Western Ghats.Their mantle-normalized element patterns show modest peaks at Rb, Ba and Pb and rather low Nb and Ta relative to La,indicating that they have been contaminated to intermediate degrees. The mantle-normalized element patterns of all the flowsand dikes show enrichment in the light rare-earth elements, with small or no Eu anomalies. The entire flow-dike sequence issimilar to the Wai Subgroup of the Western Ghats, in terms of its elemental and isotopic chemistry and stratigraphic

Žrelationships. Wai Subgroup-like lavas i.e., some of the younger magma types originally identified from the southern part of.the Western Ghats are previously known from the central, northern and northeastern Deccan, and many have been thought

to be far-travelled flows erupted in the southwestern Deccan. Although at least the dikes, and probably the giant plagioclasebasalt flows of our study area, are locally generated and emplaced, our new data extend the known outcrop area of these

112

widespread magma types substantially, and these magma types indeed appear to have a nearly province-wide distribution.

Keywords: Deccan; basalt; Tapi; rift; isotopes; geochemistry

1. Introduction

The Deccan flood basalt province covers today anarea of 500,000 km2 in western and central IndiaŽ .Fig. 1 after erosion and excluding the substantialarea downfaulted into the Arabian Sea to the westŽ .e.g., Mahoney, 1988 . The combined stratigraphicthickness of these lavas in the well-studied Western

Ž .Ghats region boxed area in Fig. 1 is ;3000 m; nostatistically significant age difference exists betweenlavas from the base and the top of this sequence, and

Žits formal mean age is 67.5"0.3 Ma Duncan and.Pyle, 1988 . This immense volcanic outburst is gen-

erally ascribed to the birth of the Reunion hotspot´beneath the northerly drifting Indian continent in the

Ž .Late Cretaceous e.g., Morgan, 1981 . The strati-graphic framework of the Western Ghats region isknown from detailed field, chemical, isotopic and

Žpalaeomagnetic studies e.g., Cox and Hawkesworth,1985; Beane et al., 1986; Lightfoot and Haw-kesworth, 1988; Mahoney, 1988 and referencestherein; Subbarao, 1988 and references therein;

.Lightfoot et al., 1990; Peng et al., 1994 , and aŽgeological map is also available Subbarao and

.Hooper, 1988 . On the basis of geochemical charac-teristics and field markers, the Western Ghats stratig-raphy has been divided into three subgroups and 11

Ž .formations Table 1 . All these lavas are thought tohave been erupted from a group of randomly ori-ented apparent feeder dikes in a region near IgatpuriŽ .Hooper, 1990 , which is considered to be the centerof an enormous shield volcano, and the lavas arenearly horizontal with regional southerly dips of

Ž-0.58 in the region south of Igatpuri Beane et al.,.1986; Devey and Lightfoot, 1986 . North of Igatpuri,

however, the dips become slightly northerly andŽ .northeasterly Subbarao et al., 1994 .

Recent and ongoing work on the central part ofthe province to the east of Igatpuri has traced theThakurvadi, Khandala and Poladpur Formations

Ž . Ž . ŽFms. for several hundred kilometers km Sub-.barao et al., 1994, 1998; Peng, 1998 . The Poladpur,

Ambenali and Mahabaleshwar Fms. also extend intothe southeastern Deccan for hundreds of km from

Žtheir type sections in the southwest e.g., Mitchell.and Widdowson, 1991 . Lavas isotopically and

chemically resembling the Ambenali, Poladpur,Khandala and Bushe Fms. have been described in theMhow, Jabalpur and Chikaldara areas in the north-

Ž .eastern Deccan Fig. 1 , but because of their gener-ally systematically higher Pb isotopic ratios, mostappear to have been erupted from different feeder

Žvents than the Western Ghats lavas Peng et al.,.1998 .

Three major rift zones cross the Deccan province:Žthe Narmada–Tapi rift zone with the Satpura horst

.in between , the Cambay rift, and the West Coast riftŽ .belt Fig. 1 . They are marked by strong structural

disturbance and faulting, high positive gravityanomalies, high heat flow, recent earthquake epicen-ters, possible Deccan magma chambers and intrusivebodies, regional dike swarms, and varied magma-

Žtism, which have led various workers e.g., Sheth.and Chandrasekharam, 1997a,b to suggest that these

rift belts had developed their own magmatic systemsindependent of the Western Ghats.

The stratigraphic status of the Deccan basalts ofthe Tapi rift region, between the Western Ghats andthe Satpura range, has remained unknown and hardlyany information has been published on their geo-chemistry or palaeomagnetism. This region shows anextensive cover of post-Deccan alluvium of the TapiRiver, from which a few ridges of the basalts rise,with confused field relationships. These ridges are

Žeither tectonic inliers or regional dikes Ravishanker,.1987; Guha, 1995; Sheth, 1998 .

Two such ENE–WSW-trending ridges in thewestern part of the Tapi rift are made up of similarsequences of flows intruded by dikes, and the de-tailed field, chemical and isotopic studies we have

113

carried out on them form the subject of the presentŽpaper. The first ridge named here the Shahada

. Ž Xridge begins 7 km south of Shahada town 21850 N,X .74830 E; see Fig. 2a . The other, Nandarde ridge, is

seen at Nandarde village, 20 km north of ShirpurŽ X X .21820 N, 74850 E .

2. Field geology

2.1. Shahada ridge

The Shahada ridge runs parallel to the Shahada–Shirpur road for about 6 km, in an ENE–WSWdirection, parallel to the Narmada–Tapi regional tec-

Ž . Ž .Fig. 1. Map of western India, modified from Peng et al. 1994 , showing the Deccan province white , the major rift zones traversing itŽ . Ž .modified from Hooper, 1990 , and important localities mentioned in the paper. The area between Igatpuri and Amboli vertical box is theWestern Ghats region.

114

Table 1Formation stratigraphy of the Deccan basalts in the Western GhatsŽ .Subbarao and Hooper, 1988

Group Subgroup Formation

Deccan Basalt Wai DesurPanhalaMahabaleshwarAmbenaliPoladpur

Lonavala BusheKhandala

Kalsubai BhimashankarThakurvadiNeralIgatpuriJawhar

tonic trend. At the Shirpur–Dondaicha road bifurca-Ž .tion Fig. 2a , the road-cut exposes a plagioclase-

phyric basalt flow in which a large multiple basalticdike has been intruded. This dike is made up of three

Žcolumnar rows, each ;6 m wide the central row isan individual dike which has been intruded along the

.median plane of the earlier dike, splitting it in two .ŽThis central dike has been weathered to large 2–3

.m rounded boulders, which are scattered along thebase of the ridge, its slopes and top. To the east,these boulders are exposed along the ridge top for itsfull length. Many small basaltic dikes are seen allalong the ridge. The stratigraphic succession of thebasalts constituting the ridge, obtained on an east-ward, downhill traverse, is shown in Fig. 2b.

2.2. Nandarde ridge

The Nandarde ridge is similar to the Shahadaridge, and is made up of an aphyric basalt flow and a

Ž . Ž .giant plagioclase basalt GPB flow Fig. 2c . Manysmall dikes are present. They contain fragments ofthe GPB at the ridge top. This GPB flow is widelyexposed to the north of Nandarde village as well.The bouldery dike at the Shahada ridge is also seenalong this ridge.

Interestingly, the Shahada and Nandarde ridgesŽseem to lie along a single lineament dashed line in

.Fig. 2a , parallel to the regional tectonic trend andtherefore reflecting some prominent structural con-trol, although because of widespread alluvial cover,

physical continuity between the ridges is absent. Theflow stratigraphic sequences for the two ridges arealso not exactly the same.

Ž .Fig. 2. a Map of the study area showing the locations of theShahada and Nandarde ridges. N.H. 3 is National Highway 3.

ŽArea falls on Survey of India toposheet No. 46K 1:250,000. Ž . Ž .scale . b Flow stratigraphy of the Shahada ridge. c Flow

stratigraphy of the Nandarde ridge. Heights are in meters abovemean sea level.

115

3. Petrography

All samples were studied in thin section on aLeitz-Laborlux petrological microscope. The miner-

Ž .als of the bouldery dike SH43, SH50 are mainlyplagioclase and clinopyroxene, with some subhedralsieve-textured opaques, orthoclase and minor sec-ondary hornblende. These rocks have a moderately

Ž .coarse grain size 3–5 mm and subophitic texture.The anorthite contents of the plagioclases were de-

Žtermined using a three-axis universal stage and later.a JEOL energy-dispersive X-ray analyser at IITB.

Most plagioclase compositions fall towards the sodicŽ .end of the series, in the oligoclase An range.10 – 30

Some labradorite grains are also present, however.The oligoclases contain many inclusions of prismaticapatite. The clinopyroxenes have optic axial anglesŽ .2V consistent with diopsidic augite. The modalcompositions of the two dike samples are as follows:

Ž .SH43 Shahada : 43% plagioclase, 44% clinopyrox-ene, 8% opaques, 4% orthoclase and 1% secondary

Ž .hornblende; SH50 Nandarde : 51% plagioclase, 36%clinopyroxene, 9% opaques and 4% orthoclase. Be-cause of the evolved plagioclase compositions, Sheth

Ž .et al. 1997 termed these rocks dioritic gabbrosŽ .oligoclase or andesine are typical of diorites . Theirnormative composition is quartz-bearing soda-gab-bro. The other dikes and the flows appear to beordinary tholeiitic basalts in thin section, with finegrain size and porphyritic–subophitic textures. Pla-gioclase is the most common phenocryst phase.However, the dike SH51 is unique in containingnumerous clusters of well-crystallized, microscopic

Ž .orthopyroxene ferro-enstatite . Each cluster containshundreds of individual orthopyroxene grains, andsuch a phenomenon previously has not been de-scribed in any other Deccan rock except an E–W-

Žtrending, 40-km-long dike near Dhule Keshav et al.,.1998; Sheth, 1998 . The origin of these clusters is

Žpresently being investigated Chandrasekharam et.al., in prep. .

4. Geochronometry

Radiometric dating of the Shahada and Nandardebouldery dike by the 40Ar–39Ar incremental heatingtechnique at Oregon State University yielded good

plateau and isochron ages of ;66 Ma. The ShahadaŽ .ridge sample SH43 was dated at 65.6"0.5 Ma and

Ž .the Nandarde ridge sample SH50 at 65.6"0.6 MaŽ .five-step weighted mean plateau ages, 1s . Com-plete 40Ar–39Ar results with plateau age and isochronage plots can be found in the work of Sheth et al.Ž .1997 .

5. Petrochemistry

5.1. Analytical methods

Major-element compositions of selected flow anddike samples, determined by inductively coupled

Ž .plasma-atomic emission spectrometry ICP-AES atthe Regional Sophisticated Instrumentation Center,IITB, are reported in Table 2. Trace-element data,also reported in Table 2, were obtained using induc-

Ž .tively coupled plasma-mass spectrometry ICP-MSat the University of Hawaii, where the isotope dilu-tion and isotope ratio measurements were also per-formed. Sample preparation and mass spectrometrictechniques for measurement of Pb, Nd and Sr iso-topic ratios and isotope dilution abundances of Pb,Nd, Sm, Sr and Rb are closely similar to methods

Ž .described by Mahoney et al. 1991 . The results,standard reference values and estimated analyticaluncertainties are presented in Table 3. ICP-AESmajor and trace element data for the dioritic gabbro

Ž .bouldery dike samples SH43, SH50 can be foundŽ .in the work of Sheth et al. 1997 .

5.2. Discussion

The isotopic ratios for four dike samples, includ-Žing the bouldery dike samples SH43, SH50, SH41

.and SH51, where ts66 Ma , are closely similar toŽ .each other. Their ´ t values are slightly positiveNd

Ž . Ž . Ž87 86q1.4 to q1.7 , and on plots of ´ t vs. SrrNd. Ž . 206 204Sr and ´ t vs. present-day Pbr Pb, datat Nd

for these dikes plot within the isotopic arrays of theMahabaleshwar and Panhala Fms. of the Western

Ž .Ghats Fig. 3a and b . In addition, plots of present-Ž .day Pb isotopic ratios Fig. 4a–c show that the data

for the dikes plot precisely within the Mahabalesh-war Fm. field.

116

Table 2Major- and trace-element compositions of the Shahada ridge–Nandarde ridge basaltsŽ . Ž .1 Major-element data were measured by ICP-AES at IITB. Data for KC-11 standard provide an idea of accuracy. 2 Mg No. is molecular,

w Ž .x Ž .assuming that 85% of the total iron is in the FeO form; thus Mg No.swt.% MgOr MgOq0.85FeO T =100. 3 LOI is percentage lossŽ .in weight of sample powder on ignition at 9008C for ;2 h. 4 Trace elements were measured by ICP-MS at UH. Data for the USGS

Ž .standard BHVO-1 provide an idea of accuracy. The reference values for KC-11 are those recommended by Walsh 1982 , and the BHVO-1Ž .reference values are from a literature-based compilation mainly Govindaraju, 1989 .

Dike Flow Flow GPB Dike Dike GPB Reference Measured

SH41 SH44 SH45 SH47 SH49 SH51 SH52 KC-11 KC-11

wt.%SiO 51.88 49.67 50.24 48.22 52.29 50.56 48.70 55.55 54.312

TiO 3.47 1.82 2.47 2.96 0.62 3.15 3.25 1.10 1.242

Al O 13.21 16.06 14.17 17.41 16.73 14.14 10.90 16.51 15.292 3Ž .FeO T 14.15 10.67 12.57 12.86 9.23 14.49 15.11 8.71 10.45

MnO 0.23 0.18 0.20 0.19 0.19 0.24 0.25 0.14 0.15MgO 5.10 5.48 5.66 4.43 7.15 6.20 4.23 3.99 4.56CaO 8.20 8.95 8.23 8.64 10.38 8.38 9.26 6.86 7.33Na O 1.52 1.95 2.56 2.36 1.61 2.13 3.11 3.30 3.512

K O 0.36 0.13 0.74 0.42 0.49 0.63 1.58 2.15 2.012

P O 0.40 0.21 0.27 0.30 0.13 0.45 0.30 0.32 0.292 5

LOI 1.64 3.70 4.52 3.08 1.74 0.50 2.18Total 100.16 98.82 101.63 100.87 100.56 100.37 98.87Mg No. 43 51 48 42 62 47 37

ppm BHVO-1 BHVO-1Co 46 42 44 42 46 51 49 45 46Ni 54 61 45 69 105 101 104 121 123Sc 41 28 43 50 37 39 31 32 30V 388 283 327 352 222 363 342 317 317Ba 192 233 205 158 206 266 285 133 137Y 49.9 22.5 29.8 34.1 24.3 43 42 27 24Cu 241 144 191 290 101 239 693 136 133Ga 30.3 30.8 27.0 30.6 22.9 33.3 31.9 21 26Rb 14 9 24 11 15.6 13 52 9.7 8.6Nb 21.2 10.3 13.7 20.6 5.4 16.3 14.8 19 17Ta 1.21 0.7 0.7 1.32 0.28 0.98 0.80 1.16 1.11Th 1.69 1.52 1.56 1.94 4.35 2.48 2.10 1.08 1.11U 0.29 0.19 0.27 0.25 1.15 0.46 0.49 0.41 0.39Pb 3.20 2.05 2.60 2.60 5.85 4.24 3.20 2.05 2.09La 21.9 15.2 20.2 21.1 15.3 25.1 21.5 15.8 15.9Ce 53.5 33.0 46.6 45.5 30.6 55.8 50.6 37.8 38.3Pr 7.22 4.55 6.08 6.3 3.5 7.57 6.70 5.40 5.38Nd 32.9 20.3 26.3 28.2 13.0 33.1 30.0 24.8 24.6Sm 8.75 4.88 6.72 6.85 3.19 8.17 7.63 6.10 5.72Eu 2.34 1.61 1.94 2.20 0.82 2.33 2.33 1.98 1.89Gd 9.81 5.43 7.02 7.50 3.46 9.10 8.73 6.40 6.06Tb 1.55 0.78 1.05 1.21 0.59 1.41 1.35 0.96 0.89Dy 8.76 4.48 5.54 6.24 3.78 8.20 7.45 5.20 5.12Ho 1.75 0.84 1.06 1.24 0.77 1.56 1.52 0.99 0.91Er 4.27 1.99 2.39 3.14 2.19 3.72 3.67 2.40 2.36Tm 0.61 0.26 0.33 0.41 0.33 0.54 0.54 0.33 0.29Yb 3.78 1.79 2.13 2.48 2.16 3.57 3.26 2.02 1.89Lu 0.57 0.28 0.28 0.37 0.32 0.50 0.46 0.29 0.26

117

Table 3Isotope dilution trace-element abundances and Nd–Sr–Pb isotopic dataŽ .1 Rb, Nd, Sm, Pb and Sr abundances were measured by isotope dilution. Analytical uncertainty on Nd and Sm abundances is -0.2%, on

ŽSr-0.5%, on Rb;1%, and on Pb-1%. Small acid-washed chips of the samples were used for this as part of the procedure for isotope.ratio determinations; e.g., see Peng and Mahoney, 1995 , and hence, the isotope dilution abundance values are not strictly representative of

the bulk rocks, although they are close to the bulk rock values.Ž . Ž . Ž87 86 .2 ´ t and Srr Sr are the initial ratios age-corrected to 66 Ma. Pb isotopic ratios are present-day values.Nd tŽ . 87 86 206 204 207 2043 Analytical uncertainty on ´ is better than "0.2 units; on Srr Sr better than 0.00002; on Pbr Pb and Pbr Pb better thanNd

"0.012, and on 208 Pbr204 Pb better than "0.038.Ž . Ž .4 Data are reported relative to the following standard values: for NBS981 Pb, to the values of Todt et al. 1984 ; for NBS987 Sr,87Srr86 Srs0.71024"0.00002; for La Jolla Nd, 143Ndr144 Nds0.511845"0.000010.

Sample Dike Dike Flow Flow Dike Dike Dike Flow

SH41 SH43 SH44 SH45 SH49 SH50 SH51 SH52

Rb ppm 13.0 23.5 7.07 20.2 14.9 34.2 13.5 55.6Nd ppm 33.76 32.03 19.55 22.83 13.20 71.14 33.98 24.28Sm ppm 8.794 8.305 4.723 5.558 2.927 10.96 8.822 6.402Pb ppm 3.33 2.74 2.09 2.07 5.63 2.75 3.42 3.14Sr ppm 206.5 212.6 300.4 328.8 142.2 230.5 210.7 261.4

Ž .´ t q1.5 q1.4 y3.9 y3.7 y20.2 q1.5 q1.7 y2.1Nd87 86Ž .Srr Sr 0.70474 0.70481 0.70782 0.70747 0.72315 0.70474 0.70471 0.70697t206 204Pbr Pb 17.595 17.483 18.423 18.485 20.935 17.578 17.607 18.506207 204Pbr Pb 15.395 15.366 15.527 15.495 16.032 15.379 15.394 15.567208 204Pbr Pb 38.227 38.144 38.874 38.923 41.401 38.326 38.208 39.331

On the other hand, isotopic data points for theŽ .flows SH44, SH45 and SH52 GPB all fall within

Ž .the Poladpur Fm. field Figs. 3 and 4 . The dikeSH49, which intrudes a lower, unanalyzed flowŽ . Ž .SH48, Fig. 2c has ´ t sy20.2 and a corre-Nd

Ž87 86 .sponding Srr Sr s0.72315; these values aret

the lowest and highest, respectively, yet known forany Deccan rock. The Bushe Fm. of the WesternGhats is the most crustally contaminated formationknown in the Deccan, and the dike SH49 has astrong Bushe Fm. affinity, but has Nd–Sr isotopicvalues even more extreme than known Bushe Fm.lavas. Interestingly, this dike has the highest Mg No.Ž .62 of all samples in this study; the combination ofhigh Mg No. and high Sr isotopic ratios has beennoted previously among lavas of the Bushe Fm.Že.g., Mahoney et al., 1982; Cox and Hawkesworth,1984, 1985; Lightfoot and Hawkesworth, 1988;

.Lightfoot et al., 1990 , and has been explained onŽ .the premise that more primitive and thus hotter

mantle magmas can experience greater amounts ofcrustal contamination during ascent than do more

Ževolved, cooler magmas e.g., O’Hara, 1978; Patch-ett, 1980; Huppert and Sparks, 1985; Mahoney,

.1988 . Significant crustal contamination has thus

probably caused the highly radiogenic Sr isotopicand unradiogenic Nd isotopic composition of this

Ždike note also its relatively high SiO with very low2Ž .TiO and FeO T , features typical of Bushe lavas2

.and of many other crustally contaminated basalts .Ž .Huppert and Sparks 1985 modelled magma ascent

in dikes, and noted that at low flow rates in narrowdikes, flow conditions should be laminar, and hence,magma should solidify against dike walls, preventingsubsequent batches from contamination. However,basaltic magmas can ascend turbulently if flow ratesare sufficiently high, and under turbulent conditions,the products of partial melting of the country rocksare continuously swept away by the magma, thusallowing further contamination to continue. The dikewidth above which magma flow is expected to beturbulent is ;3 m. Note that the outcrop width ofthe SH49 dike in the field is ;15 m. This E–W-

Ž .trending dike, and the flow it intrudes SH48 , areŽ .found 4 km south of Boradi Fig. 2a , just to the

eastern side of the Shirpur–Boradi road. They arenot seen at the Nandarde ridge itself.

Fig. 5 shows the primitive-mantle-normalizedtrace element patterns of the Shahada and Nandarderidge flows and dikes. The patterns of the isotopi-

118

Ž . Ž .Fig. 3. a Nd–Sr and b Nd–Pb isotopic plots for the Shahada and Nandarde ridge flows and dikes. Formation fields are after Peng et al.Ž .1994 .

cally Mahabaleshwar-like dikes SH41 and SH51 havemodest Pb and Ba peaks and are enriched in the lightrare-earth elements relative to the heavy ones, withsmall negative Eu anomalies resulting from plagio-

Ž .clase fractionation Fig. 5a . The pattern for theisotopically Bushe-like dike SH49 shows large peaksat Pb, Th and U, and a large trough at Nb–Ta, asexpected from crustal contamination, and is light

119

Žrare-earth-enriched with a negative Eu anomaly Fig.. Ž .5a . The patterns for the flows Fig. 5b have peaks

at Pb, Rb and Ba, troughs at Nb–Ta, and lightrare-earth enrichment with very small Eu troughs.These flows appear to have been contaminated tointermediate degrees, just like the Poladpur Fm.lavas of the Western Ghats.

6. A regional perspective

To briefly restate our main results, most dikes ofthe Shahada and Nandarde ridges are isotopicallyand chemically similar to lavas of the Mahabalesh-

Žwar Fm. of the Western Ghats southwestern Dec-.can , and the flows to the Poladpur Fm. The whole

assemblage of rocks is therefore similar to the WaiSubgroup. The presence of Wai Subgroup lavas,dominantly belonging to the Poladpur Fm., has alsobeen reported in the Buldana and Lonar areas of the

Ž .central Deccan Fig. 1 , 150–200 km SSE of ourŽarea, Subbarao et al., 1994, 1996, 1998; Peng,

.1998 .Lavas isotopically and chemically indistinguish-

able from the Ambenali Fm. constitute the top partsof sequences south of Jabalpur and north of

Ž .Chikaldara in the northeastern Deccan Fig. 1 . Theyare underlain by flows chemically resembling thePoladpur Fm., and at Chikaldara, several flows simi-lar to the Khandala Fm. Lavas with broadly Polad-pur- and Khandala-like elemental compositions arealso abundant in a thick section to the south ofMhow, and thus appear to be widespread across the

Ž .northeastern Deccan Peng, 1998; Peng et al., 1998 .However, most of the northeastern Poladpur- andKhandala-like basalts are different from the chemi-cally similar southwestern basalts in having consis-tently higher 206 Pbr204 Pb. As the general strati-graphic order in the Chikaldara and Jabalpur areas iscrudely the same as that in the southwestern DeccanŽi.e., Ambenali-like lavas overlying chemically Po-ladpur-like lavas, which in turn overlie chemically

. Ž .Khandala-like lavas , Peng et al. 1998 argued thatthey are petrogenetically related to the southwesternDeccan basalts but erupted from different feedervents; that is, the northeastern basalts are not justfar-travelled lavas erupted from the same dikes asthe southwestern ones. The locations of the feeder

dikes for the northeastern flows are unknown; manymay be relatively near the sections studied. Thus,overall, Wai Subgroup-like lavas seem to bewidespread over the central and northern Deccan,and are by no means confined to the southern regionof the Western Ghats.

However, with regards to the basalts of our studyarea, their chemical and isotopic similarities with theWai Subgroup basalts do not necessarily mean that

Žthe entire Wai Subgroup as defined from the West-.ern Ghats physically extends into our area. This

statement derives from the following observations.Ž . Ž1 Several GPBs are present in our area our unpub-lished data; SH47 and SH52 are only two examples;

.Sheth, 1998 . In the Western Ghats, GPBs are con-Ž .fined to the oldest Kalsubai subgroup only, and are

Žconspicuously absent from the younger Lonavala. Ž .and Wai subgroups Hooper et al., 1988 . Therefore,

the widespread occurrence of GPBs in the Wai Sub-group-like sequences in the Shahada–Nandarde re-gion strongly suggests that this region has had anevolutionary history at least partly independent ofthe Western Ghats. Highly plagioclase-phyric flowsand GPBs are thought to have had relatively localeruptive vents because of their expected high viscos-

Ž . Ž .ity e.g., Mahoney, 1988 . 2 The relatively dike-richNarmada–Satpura–Tapi region could have been animportant source area for the Deccan eruptions, not-ing the important role of rifting and lithospheric

Žextension e.g., Sheth and Chandrasekharam,. Ž .1997a,b . 3 A few cases of dikes and sills passing

into flows have actually been demonstrated or in-ferred for the Narmada–Satpura region, based onfield, geochemical, palaeomagnetic and geochrono-

Žmetric studies e.g., Crookshank, 1936; Subbarao etal., 1988; Bhattacharji et al., 1994; Sen and Cohen,

.1994 .The present study, supplemented by some recent

Žstudies Deshmukh et al., 1996; Yedekar et al., 1996;.Peng et al., 1998 , strongly suggests that some of the

common magma types in the Deccan are much morewidespread than generally believed previously, andour findings extend the outcrop areas of these geo-graphically widespread and volumetrically abundantmagma types substantially. However, the AmbenaliFm., which separates the Poladpur and Mahabalesh-war Fms. in the Western Ghats, is completely absentin our study area, and around Mhow, but Ambenali-

120

121

Ž . Ž .Fig. 5. Primitive-mantle-normalized incompatible trace-element patterns of a dikes and b flows from the Shahada and Nandarde ridges.Ž .Normalizing values are from Sun and McDonough 1989 .

type lavas form the top parts of more distant sectionsin the northeastern Deccan, and may have flowed

Žconsistently in a northeastern direction Peng et al.,.1998 .

7. Conclusions

Flows and dikes of the Shahada–Shirpur regionof the Tapi rift in the Deccan flood basalt province

are closely similar to the Wai Subgroup of theWestern Ghats, in terms of their elemental and iso-topic chemistry. Wai Subgroup-like lavas are previ-ously known from the central, northern and north-eastern Deccan, and many have been thought to befar-travelled flows erupted along the Western Ghats.On the other hand, at least the dikes, and probablyalso the GPB flows of our study area have beenemplaced from relatively local feeders distributedalong this rift zone. However, our new data extend

Ž . 207 204 206 204 Ž . 208 204 207 204 Ž . 208 204 206 204Fig. 4. Plots of a Pbr Pb vs. Pbr Pb, b Pbr Pb vs. Pbr Pb, and c Pbr Pb vs. Pbr Pb, for the ShahadaŽ . Ž . Ž . Ž . Ž .and Nandarde ridge flows and dikes. Formation fields in a and b are after Mahoney 1988 , and those in c are after Peng et al. 1994 .

122

the known outcrop area of the many widespreadDeccan magma types substantially, and these magmatypes indeed appear to have a nearly province-widedistribution.

Acknowledgements

Thanks are due to Dr. Surendra Pal Verma for hiskind invitation to contribute this paper. We alsoacknowledge insightful reviews from Peter Hooper,Ray Kent and Wendy Bohrson, and financial supportfrom the Department of Science and Technology,

Ž .Govt. of India to DC and HCS , and the U.S.Ž .National Science Foundation to JM .

References

Beane, J.E., Turner, C.A., Hooper, P.R., Subbarao, K.V., Walsh,J.N., 1986. Stratigraphy, composition and form of the Deccanbasalts, Western Ghats, India. Bull. Volcanol. 48, 61–83.

Bhattacharji, S., Chatterjee, N., Wampler, J.M., Gazi, M., 1994.Mafic dikes in Deccan volcanics-indicator of India intraplaterifting, crustal extension and Deccan flood basalt volcanism.

Ž .In: Subbarao, K.V. Ed. , Volcanism. Wiley Eastern, pp.253–276.

Chandrasekharam, D., Vaselli, O., Sheth, H.C., Keshav, S., inprep. Petrogenetic significance of orthopyroxene crystals inthe Tapi rift dikes.

Cox, K.G., Hawkesworth, C.J., 1984. Relative contribution ofcrust and mantle to flood basalt magmatism, Mahabaleshwararea, Deccan Traps. Philos. Trans. R. Soc. London, Ser. A310, 627–641.

Cox, K.G., Hawkesworth, C.J., 1985. Geochemical stratigraphy ofthe Deccan Traps at Mahabaleshwar, Western Ghats, India,with implications for open system magmatic processes. J.Petrol. 26, 355–377.

Crookshank, H., 1936. Geology of the northern slopes of theSatpuras between Morand and Sher Rivers. Mem. Geol. Surv.Ind. 66.

Deshmukh, S.S., Sano, T., Nair, K.K.K., 1996. Geology andchemical stratigraphy of the Deccan basalts of Chikaldara andBehramghat sections from the eastern part of the Deccan Trap

Ž .province, India. In: Deshmukh, S.S., Nair, K.K.K. Eds. ,Deccan Basalts. Gondwana Geol. Mag. Spl., Vol. 2, pp. 1–22.

Devey, C.W., Lightfoot, P.C., 1986. Volcanological and tectoniccontrol of stratigraphy and structure in the western DeccanTraps. Bull. Volcanol. 48, 195–207.

Duncan, R.A., Pyle, D.G., 1988. Rapid eruption of the Deccanflood basalts at the CretaceousrTertiary boundary. Nature333, 841–843.

Govindaraju, K., 1989. Compilation of working values and sampledescriptions for 272 geostandards. Geostand. Newsl. 13, 1–113.

Guha, S., 1995. Tectonic framework and evolution of the Tapibasin — an intracratonic half-graben from west-central India.Ind. Miner. 49, 61–78.

Hooper, P.R., 1990. The timing of crustal extension and theeruption of continental flood basalts. Nature 345, 246–249.

Hooper, P.R., Subbarao, K.V., Beane, J.E., 1988. The giantŽ .plagioclase basalts GPBs of the Western Ghats, Deccan

Ž .Traps. In: Subbarao, K.V. Eds. , Deccan Flood Basalts. Mem.Geol. Soc. Ind., 10, pp. 135–144.

Huppert, H.E., Sparks, R.S.J., 1985. Cooling and contamination ofmafic and ultramafic magmas during ascent through the conti-nental crust. Earth Planet. Sci. Lett. 74, 371–386.

Keshav, S., Sheth, H.C., Chandrasekharam, D., 1998. Field geol-ogy, petrography, and orthopyroxene clusters of the Dhule-Parola dike, Tapi valley, central Deccan basalt province. Curr.Sci. 74, 252–254.

Lightfoot, P., Hawkesworth, C., 1988. Origin of Deccan Traplavas: evidence from combined trace element and Sr-, Nd- andPb-isotopic studies. Earth Planet. Sci. Lett. 91, 89–104.

Lightfoot, P.C., Hawkesworth, C.J., Devey, C.W., Rogers, N.W.,van Calsteren, P.W.C., 1990. Source and differentiation ofDeccan Trap lavas: implications of geochemical and mineralchemical variations. J. Petrol. 31, 1165–1200.

Ž .Mahoney, J.J., 1988. Deccan Traps. In: Macdougall, J.D. Ed. ,Continental Flood Basalts. Kluwer Academic Publishers, Dor-drecht, The Netherlands, pp. 151–194.

Mahoney, J.J., Macdougall, J.D., Lugmair, G.W., Murali, A.V.,Sankar Das, M., Gopalan, K., 1982. Origin of the Deccan Trapflows at Mahabaleshwar inferred from Nd and Sr isotopic andchemical evidence. Earth Planet. Sci. Lett. 60, 47–60.

Mahoney, J.J., Nicollet, C., Dupuy, C., 1991. Madagascar basalts:tracking oceanic and continental sources. Earth Planet. Sci.Lett. 104, 350–363.

Mitchell, C., Widdowson, M., 1991. A geological map of thesouthern Deccan Traps, India and its structural implications. J.

Ž .Geol. Soc. London 148, 495–505.Morgan, W.J., 1981. Hotspot tracks and the opening of the

Ž .Atlantic and Indian Oceans. In: Emiliani, C. Ed. , The Sea,Vol. 7. Wiley, New York, pp. 443–487.

O’Hara, M.J., 1978. Thermal history of magmas: the low pressurereference point. Philos. Trans. R. Soc. London, Ser. A 288,627–629.

Patchett, P.J., 1980. Thermal effects of basalt on continental crustand crustal contamination of magmas. Nature 283, 559–561.

Peng, Z.X., 1998. Geochemical studies of the northwestern, cen-tral and northeastern Deccan Traps, India: sources, petrogene-sis, and relations to southwestern Deccan stratigraphy. PhDdissertation, University of Hawaii, USA.

Peng, Z.X., Mahoney, J.J., 1995. Drillhole lavas from the north-western Deccan Traps, and the evolution of Reunion hotspot´mantle. Earth Planet. Sci. Lett. 134, 169–185.

Peng, Z.X., Mahoney, J., Hooper, P., Harris, C., Beane, J., 1994.A role for lower continental crust in flood basalt genesis?Isotopic and incompatible element study of the lower six

123

formations of the western Deccan Traps. Geochim. Cos-mochim. Acta 58, 267–288.

Peng, Z.X., Mahoney, J.J., Hooper, P.R., Macdougall, J.D., Krish-namurthy, P., 1998. Basalts of the northeastern Deccan Traps,India: isotopic and elemental geochemistry and relation tosouthwestern Deccan stratigraphy. J. Geophys. Res., in press.

Ravishanker, 1987. Neotectonic activity along the Tapi–Satpuralineament in central India. Ind. Miner. 41, 19–30.

Sen, G., Cohen, T.H., 1994. Deccan intrusion, crustal extension,doming, and the size of the Deccan-Reunion plume head. In:´

Ž .Subbarao, K.V. Ed. , Volcanism. Wiley Eastern, pp. 201–216.Sheth, H.C., 1998. Geochemistry, petrogenesis, stratigraphy and

structure of Deccan flood basalts of the western Satpura–Tapiregion, India. PhD dissertation, Indian Institute of Technology,Bombay, India.

Sheth, H.C., Chandrasekharam, D., 1997a. Plume–rift interactionin the Deccan volcanic province. Phys. Earth Planet. Inter. 99,179–187.

Sheth, H.C., Chandrasekharam, D., 1997b. Early alkaline magma-tism in the Deccan Traps: implications for plume incubationand lithospheric rifting. Phys. Earth Planet. Inter. 104, 371–376.

Sheth, H.C., Duncan, R.A., Chandrasekharam, D., Mahoney, J.J.,1997. Deccan Trap dioritic gabbros from the westernSatpura–Tapi region. Curr. Sci. 72, 755–757.

Ž .Subbarao, K.V. Ed. , 1988. Deccan Flood Basalts. Mem. Geol.Soc. Ind., 10, xxivq393 pp.

Subbarao, K.V., Hooper, P.R., 1988. Reconnaissance map of theDeccan Basalt Group in the Western Ghats, India. In: Sub-

Ž .barao, K.V. Ed. , Deccan Flood Basalts. Mem. Geol. Soc.Ž .Ind., 10 enclosure .

Subbarao, K.V., Ramasubba Reddy, N., Prasad, C.V.R.K., 1988.Geochemistry and palaeomagnetism of dykes from Man-

daleswar region, Deccan basalt province. In: Subbarao, K.V.Ž .Ed. , Deccan Flood Basalts. Mem. Geol. Soc. Ind., 10, pp.225–233.

Subbarao, K.V., Chandrasekharam, D., Navaneethakrishnan, P.,Hooper, P.R., 1994. Stratigraphy and structure of parts of thecentral Deccan basalt province: eruptive models. In: Subbarao,

Ž .K.V. Ed. , Volcanism. Wiley Eastern, pp. 321–332.Subbarao, K.V., Peng, Z.X., Mahoney, J.J., Hooper, P.R., Nava-

neethakrishnan, P., 1996. Current studies on the geochemistryof the Deccan Traps from the central region. Deccan ’96 Intl.Conf. Abstracts, 31.

Subbarao, K.V., Peng, Z.X., Mahoney, J.J., Hooper, P.R., Nava-neethakrishnan, P., 1998. Chemical and isotopic stratigraphyof the Central Deccan basalt province with implications for

Ž .regional structure. In: Subbarao, K.V. Ed. , Mem. Geol. Soc.Ind., submitted.

Sun, S.-s., McDonough, W.F., 1989. Chemical and isotopic sys-tematics of oceanic basalts: implications for mantle composi-

Ž .tion and processes. In: Saunders, A.D., Norry, M.J. Eds. ,Magmatism in the Ocean Basins. Geol. Soc. Spec. Publ., 42,pp. 313–345.

Todt, W., Cliff, R.A., Hanser, A., Hofmann, A.W., 1984. 202 Pbq205Pb double spike for lead isotopic analyses. Terra Cognita4, 209.

Walsh, J.N., 1982. Whole rock analysis by inductively coupledplasma spectrometry. Inst. Min. Metall., pp. 79–91.

Yedekar, D.B., Aramaki, S., Fujii, T., Sano, T., 1996. Geochemi-cal signature and stratigraphy of the Chhindwara–Jabalpur–Seoni–Mandla sector of the Eastern Deccan volcanic provinceand problems of its correlation. In: Deshmukh, S.S., Nair,

Ž .K.K.K. Eds. , Deccan Basalts. Gondwana Geol. Mag. Spl.,Vol. 2, pp. 49–68.