Provenance And Weathering Of Sandstones

GEOCHEMISTRY OF SANDSTONES FROM THE UPPER MIOCENE KUDANKULAM FORMATION, SOUTHERN INDIA: IMPLICATIONS FOR PROVENANCE, WEATHERING, AND TECTONIC SETTING J.S. ARMSTRONG-ALTRIN,1 YONG IL LEE, 2 SURENDRA P. VERMA,1 AND S. RAMASAMY3 1

Centro de Investigacio´n en Energı´a, Universidad Nacional Auto´noma de Me´xico (UNAM), Priv. Xochicalco S/No., Col. Centro, Apartado Postal 34, Temixco, Morelos 62580, Me´xico e-mail: [email protected] and [email protected] 2 School of Earth and Environmental Sciences, Seoul National University, Seoul 151-747, South Korea 3 Department of Geology, University of Madras, Guindy Campus, Chennai 600 025, India

ABSTRACT: Petrographic, major, trace, and rare earth element compositions of sandstones from the upper Miocene Kudankulam Formation, Southern India, have been investigated to determine their provenance, tectonic setting, and weathering conditions. All sandstone samples are highly enriched in quartz (Q) but poor in feldspar (F) and lithic fragments (L). The major-element concentrations of these sandstones reveal the relative homogeneity of their source. Geochemically, the Kudankulam sandstones are classified as arkose, subarkose, litharenite, and sublitharenite. The CIA values (chemical index of alteration; mean value ; 44.5) for these sandstones and the A–CN–K diagram suggest their low-weathering nature. Similarly, their Fe2O3* 1 MgO (mean ; 2.7), Al2O3/SiO2 (; 0.09), K2O/Na2O (; 2.2) ratios and TiO2 contents (; 0.3) are consistent with a passive-margin setting. The Eu/Eu* (; 0.5), (La/Lu)cn (; 21), La/Sc (; 5.9), Th/Sc (; 1.9), La/Co (; 5.7), Th/Co (; 1.8), and Cr/Th (; 5.3) ratios support a felsic source for these sandstones. Chondrite-normalized REE patterns with LREE enrichment, flat HREE, and negative Eu anomaly also are attributed to felsic source-rock characteristics for Kudankulam sandstones. Total REE concentrations of these sandstones reflect the variations in their grain-size fractions. The source rocks are probably identified to be Proterozoic gneisses, charnockites, and granites of the Kerala Khondalite Belt, which must have been exposed at least since the late Miocene. Finally, the unusual Ni enrichment in the Kudankulam sandstones, unaccompanied by a similar enrichment in Cr, Co, and V, may be related to either the presence of pyrite in the sandstones or, more likely, the fractionation of garnet from the source rocks during transportation.

INTRODUCTION

Sedimentary rocks are principal sources of information concerning past conditions on the Earth’s surface. Clastic rocks may preserve detritus from long-eroded source rocks and may provide the only available clues to the composition and timing of exposure of such source rocks. Geochemistry of sedimentary rocks may complement the petrographic data, especially when the latter are ambiguous. The geochemical composition of sedimentary rocks is a complex function of various variables such as source material, weathering, transportation, physical sorting, and diagenesis (Middleton 1960; Piper 1974; Bhatia 1983; McLennan 1989; Cox and Lowe 1995). Examples of using geochemical data from sediments for understanding sedimentary processes such as weathering, provenance, diagenesis, sorting, and recycling are increasing in the literature because of the sensitiveness of some key trace elements in identifying minor components that are not readily recognized petrographically (e.g., Hiscott 1984; Garver et al. 1996). Several trace elements, such as the rare earth elements (REE; e.g., La, Ce, Nd, Gd, Yb), Y, Th, Zr, Hf, Nb, and Sc are most suited for discriminations of provenance and tectonic setting because of their relatively low mobility during sedimentary processes and their short residence times in seawater (Holland 1978; Taylor and McLennan 1985). These elements probably are transferred quantitatively into clastic sediments during weathering and transportation, reflecting the signature of the parent materials, JOURNAL OF SEDIMENTARY RESEARCH, VOL. 74, NO. 2, MARCH, 2004, P. 285–297 Copyright q 2004, SEPM (Society for Sedimentary Geology) 1527-1404/04/074-285/$03.00

and hence are expected to be more useful in discriminating tectonic environments and source-rock compositions than the major elements (Bhatia and Crook 1986; McLennan 1989; Condie 1993). Rocks of southern India are older than 2500 Ma, and they are regarded to have been first exposed to the surface possibly during the Tertiary and, at places, as late as the Quaternary (Singh and Rajamani 2001a, 2001b). Geochemical characterisitics of Archean terranes are likely to be substantially different than in younger environments. For example, granitic rocks formed during the Archean are more commonly Na- and plagioclase-rich granodiorites–tonalites, whereas during the Phanerozoic they tend to be more K- and K-feldspar-rich granodiorites–monzonites–granites (Taylor and McLennan 1985; Goodwin 1991). Thus, sedimentary rocks of any age, derived primarily from Precambrian terranes, may be influenced by these differences. On the basis of geochemistry of flood-plain sediments of the Cauveri River, southern India (Fig. 1A), Singh and Rajamani (2001a) interpreted that exhumation of Archean deep crustal rocks occurred in geologically recent times. The purpose of this study is to identify the provenance and to test the neotectonic activity in southern India by examining geochemisty of upper Miocene–Pliocene sandstones cropping out in the southern tip of the Indian Peninsula. To know the probable source rock for the Kudankulam sandstones we compared our data with the gneisses, charnockites, and granites of Proterozoic Kerala Khondalite Belt, southern India (Chacko et al. 1992; Braun et al. 1996), which is located very near to our study area (see the rectangle marked Kerala Khondalite Belt in Fig. 1A). In addition, the geochemical features of clastic sedimentary rocks reflect not only the nature and proportion of detrital components but also the chemical characterisitcs of authigenic minerals formed during diagenesis. Understanding the relative importance of these factors through geochemical tools extends the knowledge of the processes that produced and affected rock deposition and lithification (e.g., Fralick and Kronberg 1997 and references therein). Enriched concentrations of certain trace elements such as Ni, Cr, Co, and V are very important for discrimination of provenance and tectonic setting. Some studied Kudankulam sandstones have enriched Ni values but low contents of Cr, Co, and V. Detrital and/or diagenetic control, and fractionation during transportation, on this peculiar geochemical feature will be also addressed in this paper. GEOLOGY AND STRATIGRAPHY OF THE STUDY AREA

The Kudankulam area in Tamil Nadu, South India (Fig. 1A), forms the southern extension of the Cauveri River Basin and is limited between Tiruchendur in the northeast and Kanniyakumari (Cape Comorin) in the south (Lat. 88 59 3099 N to 88 309 1299 N and Long. 778 309 1099 E to 788 109 1299 E). The lithostratigraphy of the Kudankulam Formation comprises the metamorphosed Proterozoic and Tertiary sedimentary rocks (Paramasivam and Srinivasan 1980). The older rocks include charnockites, quartzofeldspathic gneiss, and granitic gneiss, whereas the Tertiary strata comprise calcareous sandstone and fossiliferous limestone, which are overlain by subrecent to recent soil cover. The sedimentary rocks include both clastic and carbonate rocks with various faunal contents such as molluscs, bryozoans, foraminifers, and fragments of red and blue-green algae (Armstrong Altrin Sam et

286

J.S. ARMSTRONG-ALTRIN ET AL.

FIG. 1.—A) Simplified geological map of Southern India showing the location of the study area (modified after Singh and Rajamani 2001b). The rectangle marked Kerala Khondalite Belt refers to the area of probable source rocks (gneisses, charnockites, and granites of Proterozoic Kerala Khondalite Belt; Chacko et al. 1992; Braun et al. 1996) for the Kudankulam sandstones. B) Simplified geological map of the Kudankulam area, showing sample locations.

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA

287

FIG. 2.—Columnar sections of the Kudankulam Formation arranged in SW to NE direction, showing sample locations. Stratigraphic section names are according to the nearest village (Fig. 1B). Blank crossed space in the Manapaud section represents the area with no exposures.

al. 2001). These Tertiary rocks were deposited directly over the Proterozoic rocks, when a considerable proportion of the land along the coast was occupied by the sea as a result of transgression during the Late Tertiary. The contact between the older and Tertiary rocks is rather sinuous, suggesting that the Tertiary rocks were deposited on a shallow continental shelf or shoreline gently dipping towards the sea. The Kudankulam Formation was correlated with Karaikal beds, of late Miocene to Pliocene age (Ramanathan 1979; Paramasivam and Srinivasan 1980). The last marine transgression in southern India occurred in the early Miocene, and its influence can be observed in Gujarat in the western part of India, and also in Kerala in the southern part (Fig. 1A; Bruckner 1988). On the eastern flank of the Cauveri Basin in Sri Lanka, the Jaffna Formation is most probably the equivalent formation to the Kudankulam Formation, on the basis of the fossil assemblages and facies characteristics. The presence of microfossils like Austritrillina howchini and Taberina malabarica places the Jaffna Formation in the upper part of the upper Miocene, specifically in the Burdigalian stage (Sahani 1979; Cooray 1984). Therefore, the Kudankulam Formation is either of Burdigalian or of late Miocene–Pliocene age. Armstrong Altrin Sam and Ramasamy (1997, 1999) discussed the petrography, stratigraphy, and depositional history of the Kudankulam Formation. The authors subdivided this formation into five lithostratigraphic units: algal limestone, sandy shell limestone, silty clay, clastic limestone, and calcareous sandstone (Fig. 1B). The facies association and the sediment composition show that the Kudankulam carbonates were deposited in a shallow marine nearshore environment (Ramasamy and Armstrong Altrin Sam 1998). A preliminary stable-isotope study on the Kudankulam limestone reveals that fresh-water circulation played a major role in diagenesis (Armstrong Altrin Sam et al. 2001).

SAMPLING AND METHODS

From eight sections in the Kudankulam area (Fig. 2) fresh rock samples were collected from outcrops exposed in stream cuts and road cuts and were washed thoroughly in distilled water to remove dust contamination. The samples were disaggregated by following the procedure adopted in Cox and Lowe (1996). Grain-size analysis was carried out in a Ro-Tap sieve shaker using American Society for Testing and Material (ASTM) sieves ranging from—1.5 f to 4.25 f at 0.50 f intervals for 20 minutes (Folk 1966). Cumulative curves were constructed to calculate the statistical grain-size parameters (mean grain size and sorting values) by applying the equations of Folk and Ward (1957). Twenty thin sections were selected for detailed petrographic study. Four hundred framework grains were counted from each thin section. Matrix and cement were not counted. The point counts were done using both Gazzi–Dickinson (Gazzi 1966; Dickinson 1970) and traditional methods. Forty-five samples were analyzed for major oxides using an analytical method adopted from Shapiro and Brannock (1962) and Shapiro (1975). Aliquots of 50 mg samples were fused with NaOH in a nickel crucible, and one aliquot, solution A, was prepared for SiO2 and Al2O3 determination. Another aliquot, solution B, was prepared by digesting the samples with HF 1 HCl 1 HNO3 acid mixture to estimate Na2O, K2O, Fe2O3*, CaO, MgO, TiO2, and MnO. Total iron, silica, alumina, titania, and manganese were determined using a Bausch and Lomb Spectronic 20 spectrophotometer. Calcium and magnesium were determined by a titration method using EDTA with screened calcite and O-cresolpthalein complexion indicator. Sodium and potassium were analyzed by using an Aimil flame photometer calibrated using standard salt solutions. For the determination of CaO in the silicate fraction, samples were separately treated with 1M

288

J.S. ARMSTRONG-ALTRIN ET AL.

TABLE 1.—Major-element concentrations in weight percent (wt. %) for sandstones of the Kudankulam Formation along with their mean grain size (MZ) and sample standard deviation (S) in f units, Chemical index of alteration (CIA, Nesbitt and Young 1982), and Plagioclase index of alteration (PIA, Fedo et al. 1995). Arkose

Rock Type Sample # MZ S SiO2 TiO2 Al2O3 Fe2O3* MnO MgO CaO Na2O K2O LOI Sum CaO* CIA PIA K2O/Al2O3 K2O/Na2O SiO2 /Al2O3 Al2O3 /SiO2 Fe2O3* 1 MgO

E4

G2

F8

A2

D4

D2

C2

F5

E9

H8

E14

E3

D3

C6

E8

D1

H15

1.02 1.96 63.98 0.38 7.54 0.55 0.01 1.20 13.61 0.53 1.55 10.01 99.36 3.94 43.70 42.18 0.21 2.93 8.49 0.12 1.96

1.03 1.81 46.88 0.24 4.12 1.21 0.20 0.75 20.81 0.78 2.20 22.75 99.76 0.72 45.31 40.15 0.53 2.82 11.38 0.09 2.55

1.08 1.76 47.20 0.33 4.62 1.06 0.03 0.58 17.71 1.02 2.15 24.96 99.66 2.05 37.40 29.79 0.47 2.11 10.22 0.10 2.20

1.26 1.74 65.80 0.45 6.08 0.58 0.01 1.05 13.50 1.02 1.68 9.50 99.67 2.79 41.50 38.70 0.28 1.65 10.82 0.09 1.81

1.38 1.85 60.40 0.32 5.91 0.99 0.01 1.65 16.40 0.56 1.90 12.10 100.24 2.63 43.23 40.32 0.32 3.39 10.22 0.10 3.00

1.64 1.78 49.80 0.25 5.66 1.08 0.03 1.27 21.50 0.58 2.09 18.00 100.26 2.47 42.34 38.43 0.37 3.60 8.80 0.11 2.86

2.00 1.55 54.20 0.32 5.21 1.10 0.03 1.04 18.70 1.01 1.71 16.21 99.53 1.77 43.63 40.77 0.33 1.69 10.40 0.10 2.57

2.06 1.39 52.47 0.40 6.12 1.77 0.03 0.78 14.62 0.40 1.65 21.78 100.02 2.69 45.49 43.85 0.27 4.13 8.57 0.12 3.26

2.08 1.40 49.79 0.14 4.78 0.51 0.02 0.91 19.92 0.78 1.26 21.90 100.01 1.61 46.16 44.79 0.26 1.62 10.42 0.10 1.82

2.25 1.45 60.72 0.27 9.32 2.31 0.02 0.66 13.46 1.20 3.20 8.75 99.91 3.04 45.95 43.84 0.34 2.67 6.52 0.15 3.26

2.41 1.37 56.81 0.33 6.98 1.08 0.01 0.56 17.52 1.05 1.51 13.79 99.64 1.98 50.07 50.07 0.22 1.44 8.14 0.12 1.91

2.47 1.20 56.21 0.26 5.12 1.21 0.02 0.71 14.50 0.74 1.78 19.85 100.40 1.89 43.76 40.70 0.35 2.41 10.98 0.09 2.38

2.63 1.35 52.74 0.35 5.38 1.53 0.01 1.05 16.16 0.56 1.78 19.95 99.51 1.73 47.30 45.92 0.33 3.18 9.80 0.10 3.24

2.65 1.00 53.73 0.17 6.97 0.82 0.03 0.64 15.21 0.39 0.81 21.45 100.22 2.89 50.12 50.14 0.12 2.08 7.71 0.13 1.85

2.75 0.90 50.98 0.36 6.75 1.65 0.03 0.61 16.58 1.30 2.76 19.14 99.89 1.34 48.67 47.67 0.41 2.68 7.55 0.13 2.80

2.86 0.94 45.33 0.18 5.64 1.36 0.02 0.88 24.77 0.75 2.25 18.86 100.04 1.29 48.39 47.24 0.40 3.00 8.04 0.12 2.76

3.00 1.01 62.29 0.48 7.00 0.81 0.03 1.21 13.52 0.47 1.60 12.94 100.35 2.41 50.41 50.54 0.23 3.40 8.90 0.11 2.31

Arkose

Rock Type E6

H10

C3

Mean (n 5 20)

E1

E2

F6

B2

B4

A1

H4

H6

F7

H2

C5

G3

3.02 0.96 62.03 0.26 9.58 2.57 0.03 0.85 15.28 1.68 2.73 5.52 100.50 2.60 47.84 46.93 0.28 1.63 6.47 0.15 3.60

3.13 0.90 59.24 0.16 8.52 2.06 0.02 0.70 13.49 1.25 3.32 11.12 99.88 1.69 49.41 48.99 0.39 2.66 6.95 0.14 3.11

3.25 0.80 41.50 0.20 6.04 0.87 0.03 0.75 27.51 0.98 2.33 19.96 100.17 1.37 47.69 46.16 0.39 2.38 6.87 0.15 2.02

2.2 6 0.8 1.36 6 0.38 55 6 7 0.29 6 0.10 6.4 6 1.5 1.3 6 0.6 0.02 6 0.01 0.89 6 0.28 17 6 4.0 0.84 6 0.33 2.0 6 0.6 16 6 6 99.95 6 0.32 2.2 6 0.8 45.9 6 3.4 44 6 5 0.32 6 0.10 2.6 6 0.8 8.9 6 1.6 0.12 6 0.21 2.6 6 0.6

1.18 1.95 48.23 0.39 3.78 0.98 0.01 1.21 21.98 1.05 2.22 19.69 99.54 0.85 39.98 29.62 0.59 2.11 12.76 0.08 2.74

1.20 1.92 58.10 0.27 4.68 1.06 0.01 1.29 19.00 0.71 1.46 13.70 100.28 2.38 39.81 36.07 0.31 2.06 12.42 0.08 2.71

1.20 1.85 42.45 0.32 1.98 0.61 0.02 1.30 30.63 0.56 0.77 21.42 100.06 0.52 42.31 38.05 0.39 1.38 21.44 0.05 2.43

1.29 1.86 43.47 0.33 1.65 0.65 0.01 1.80 31.63 0.45 0.78 19.10 99.87 0.50 39.82 32.82 0.47 1.73 26.35 0.04 3.03

1.47 1.25 42.50 0.28 2.45 1.05 0.01 0.99 30.01 0.58 1.05 21.29 100.21 0.50 44.96 41.35 0.43 1.81 17.35 0.06 2.58

1.50 1.46 55.12 0.28 4.14 1.18 0.01 0.72 14.63 1.68 2.17 19.86 99.79 0.55 40.38 32.25 0.52 1.29 13.31 0.08 2.38

1.50 1.08 48.82 0.24 3.01 2.05 0.02 0.73 20.02 1.02 1.98 21.83 99.72 0.63 37.74 23.49 0.66 1.94 16.22 0.06 3.57

1.57 1.10 44.62 0.21 2.41 1.10 0.01 0.74 29.16 0.56 1.22 20.08 100.11 0.55 42.64 36.19 0.51 2.18 18.51 0.05 2.30

1.68 1.05 56.20 0.24 4.89 0.71 0.02 1.20 18.60 0.65 1.54 16.20 100.25 1.68 45.78 43.87 0.31 2.37 11.49 0.09 2.27

1.81 0.97 53.21 0.27 3.01 0.91 0.02 0.48 15.21 1.03 2.01 24.06 100.21 0.96 34.89 19.52 0.67 1.95 17.68 0.06 1.83

1.89 1.01 56.70 0.26 4.56 1.36 0.01 0.98 16.00 0.81 2.04 17.26 99.98 1.97 39.03 32.37 0.45 2.52 12.43 0.08 2.83

2.01 0.92 48.28 0.44 2.98 0.65 0.01 0.44 19.34 0.30 0.70 27.31 100.45 1.19 46.60 45.55 0.23 2.33 16.20 0.06 1.49

Sample # MZ S SiO2 TiO2 Al2O3 Fe2O3* MnO MgO CaO Na2O K2O LOI Sum CaO* CIA PIA K2O/Al2O3 K2O/Na2O SiO2 /Al2O3 Al2O3 /SiO2 Fe2O3* 1 MgO

Subarkose

Rock Type Sample # MZ S SiO2 TiO2 Al2O3 Fe2O3* MnO MgO CaO Na2O K2O LOI Sum CaO* CIA PIA K2O/Al2O3 K2O/Na2O SiO2 /Al2O3 Al2O3 /SiO2 Fe2O3* 1 MgO

Subarkose

Litharenite

Sublitharenite

F3

C7

Mean (n 5 14)

G1

H1

F2

E13

G4

H3

B6

C4

F10

Mean (n 5 9)

A3

G5

2.74 0.86 55.30 0.36 4.25 1.24 0.02 0.71 19.10 0.86 2.23 15.83 99.90 1.04 42.63 35.71 0.52 2.59 13.01 0.08 2.32

3.05 0.81 58.00 0.33 4.68 0.91 0.03 0.65 17.30 0.69 2.01 14.90 99.50 1.14 46.51 43.84 0.43 2.91 12.39 0.08 1.84

1.7 6 0.6 1.29 6 0.43 51 6 8 0.30 6 0.06 3.5 6 1.1 1.03 6 0.37 0.02 6 0.01 0.95 6 0.38 22 6 6 0.78 6 0.34 1.6 6 0.6 19.5 6 3.7 99.99 6 0.29 1.0 6 0.6 41.7 6 3.5 35 6 8 0.46 6 0.13 2.08 6 0.45 15.8 6 4.2 0.07 6 0.01 2.4 6 0.5

1.79 1.82 53.21 0.28 5.71 2.15 0.01 1.01 16.33 0.84 1.25 19.53 100.32 1.96 47.55 46.84 0.22 1.49 9.32 0.11 3.91

2.07 1.80 50.01 0.29 4.50 1.92 0.02 1.03 15.91 1.51 0.78 23.53 99.50 1.50 42.63 41.23 0.17 0.52 11.11 0.09 3.88

2.26 1.82 55.08 0.09 5.62 1.38 0.01 0.82 18.03 0.30 0.85 17.71 99.89 2.83 46.15 45.46 0.15 2.83 9.80 0.10 2.68

2.73 1.75 53.49 0.20 5.62 1.53 0.02 0.93 18.25 0.63 1.00 18.54 100.21 2.31 47.07 46.43 0.18 1.59 9.52 0.11 3.01

2.74 1.26 61.60 0.30 6.02 2.30 0.02 0.79 14.50 0.63 1.85 11.72 99.73 2.85 42.27 39.25 0.31 2.94 10.23 0.10 3.51

2.84 1.01 55.99 0.30 5.71 1.71 0.01 0.83 14.37 0.81 1.10 18.89 99.72 1.84 49.32 49.14 0.19 1.36 9.81 0.10 3.14

2.86 0.95 54.29 0.31 5.95 1.97 0.01 0.51 17.81 0.40 0.95 18.29 100.49 2.51 48.77 48.52 0.16 2.38 9.12 0.11 3.02

3.00 0.91 59.45 0.27 7.43 2.34 0.02 0.82 13.50 0.95 1.24 14.02 100.04 2.79 48.22 47.85 0.17 1.31 8.00 0.13 3.67

3.00 0.93 56.96 0.41 6.05 1.37 0.01 1.05 13.08 0.50 1.23 19.23 99.89 2.37 48.35 47.91 0.20 2.46 9.42 0.11 3.00

2.59 6 0.44 1.36 6 0.43 55.6 6 3.5 0.27 6 0.09 5.8 6 0.8 1.8 6 0.4 0.01 6 0.01 0.87 6 0.17 15.8 6 2.0 0.73 6 0.36 1.14 6 0.32 17.9 6 3.4 99.98 6 0.32 2.3 6 0.5 46.7 6 2.6 45.9 6 3.4 0.19 6 0.05 1.9 6 0.8 9.6 6 0.8 0.10 6 0.01 3.31 6 0.44

1.26 1.80 47.37 0.15 2.35 1.61 0.02 0.80 19.38 0.77 0.75 26.51 99.71 1.02 37.40 33.01 0.32 0.97 20.16 0.05 3.29

2.02 0.93 49.10 0.18 3.13 1.68 0.01 1.41 23.98 0.54 0.93 19.10 100.06 1.16 43.87 41.47 0.30 1.72 15.69 0.06 3.82

CaO* 5 CaO in silicate phase; Fe2O3* 5 Total Fe expressed as Fe2O3.

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA

289

FIG. 4.—Geochemical classification of Kudankulam sandstones using log(SiO2 / Al2O3)—log(Fe2O3*/K2 O) diagram (after Herron 1988).

FIG. 3.—QFL diagram with tectonic fields of Dickinson and Suczek (1979). Q, total quartz (monocrystalline and polycrystalline grains); F, feldspars (plagioclase and K-feldspars); L, lithic rock fragments (excluding carbonates). Kudankulam sandstones (20 selected samples) fall entirely within the field of craton-interior sources.

cold dilute HCl acid before digestion and were analyzed separately. Our chemical analyses have precisions better than 5% for all elements determined in our samples. The major-element data were recalculated on an anhydrous (LOI-free) basis and adjusted to 100% before using them in various diagrams. Twenty-one representative samples were analyzed for trace-element and REE geochemistry at the Korea Basic Science Institute. Trace-element (Ba, Co, Cr, Cu, Ni, Sc, Sr, V, Zn, and Zr) concentrations were determined using a Jobin Yvon 138 Ultrace inductively coupled plasma atomic emission spectrometer (ICP-AES). REE and some additional trace elements (Cs, Hf, Nb, Pb, Rb, Th, U, and Y) were analyzed by a VG Elemental PQII Plus inductively coupled plasma mass spectrometer (ICP-MS) using a method given by Jarvis (1988). Analytical precision for trace elements and REE is generally better than 5%. United States Geological Standard MAG1 was used for calibration. Three analyses were made for each sample and averaged. For REE discussion we used chondrite normalization factors listed in Taylor and McLennan (1985).

are common among the opaque grains. All sandstone samples contain minor amounts of feldspar grains (mean ; 10%). Both orthoclase and plagioclase feldspars are present, but K-feldspar dominates. The rock fragments are comparatively less abundant, and consist of dominantly sedimentary rock fragments. The average quartz–feldspar–lithic fragment (QFL) ratio is Q87:F10:L3. Most of the samples are free of matrix. On a QFL diagram (Fig. 3) the Kudankulam sandstones plot in the field of a craton-interior source, indicating that they were derived from igneous source rocks (Dickinson and Suczek 1979). Calcite cement occurs in all sandstones. Three types of calcite cements are observed: micrite, microsparite, and sparry calcite. Calcite cement produced corrosion on detrital grains, particularly in quartz. Sparry calcite cement is interpreted to have been formed by the movement of ground water saturated with calcium carbonate. Major Elements The major-element concentrations of all Kudankulam sandstones are arranged in Table 1 according to rock type and decreasing mean grain size

RESULTS

Petrography A textural study was carried out for the Kudankulam sandstones to characterize grain-size variations. The mean grain size (MZ expressed in f units) of quartz grains of the sandstones ranges from 1.02 f to 3.25 f (Table 1), suggesting that the sand grains are medium to very fine in size. The standard deviation values of sandstones vary from 0.80 f (moderately sorted) to 1.96 f (poorly sorted; Table 1). The framework grains are non-undulatory monocrystalline quartz (Qn), undulatory monocrystalline quartz (Qu), polycrystalline quartz (Qp), Kfeldspar, plagioclase, and rock fragments. Quartz is the most abundant framework grain in the sandstones, constituting on average 87% of rock volume. Among quartz grains, Qn is dominant over Qu. Fluid globules and tiny gas bubbles are present in Qn as transport lines, suggesting their igneous origin. Qp is a minor constituent and exhibits straight sub-grain boundaries. Heavy minerals such as garnets are rare; ilmenite and magnetite

FIG. 5.—K2O/Na2O—SiO2 /Al2O3 bivariate plot for the Kudankulam sandstones. Sandstone samples from this study; average data for comparison are from 2Pettijohn et al. (1972); 3Chacko et al. (1992); 4Braun et al. (1996); 5Condie (1993); 6Taylor and McLennan (1985). UCC 5 upper continental crust. 1

290

J.S. ARMSTRONG-ALTRIN ET AL. TABLE 2.—Trace-element concentrations in ppm for sandstones of the Kudankulam Formation. Arkose

Rock Type

Subarkose

Sample #

E4

A2

D4

D2

C2

H8

D1

E6

H10

C3

E2

F6

B2

B4

Mz Ba Co Cr Cs Cu Hf Nb Ni Pb Rb Sc Sr Th U V Y Zn Zr

1.02 30.35 1.32 5.88 0.05 7.06 0.40 0.73 26.82 2.07 5.41 1.82 87.06 0.85 0.19 7.77 5.07 2.82 4.71

1.26 38.20 1.74 13.73 0.08 7.08 0.62 6.01 24.00 6.02 7.30 1.64 50.20 2.36 1.40 9.73 8.81 28.40 4.80

1.38 151.0 1.04 21.00 0.06 22.70 0.83 4.81 106.0 12.50 13.50 4.07 103.0 5.03 1.28 29.30 7.09 16.90 31.00

1.64 41.20 3.01 14.68 0.08 11.20 0.29 1.79 33.60 35.30 8.63 1.97 95.3 5.28 0.61 13.60 7.02 13.80 6.09

2.00 101.0 2.64 10.30 0.20 16.40 0.85 1.47 30.70 4.37 10.80 2.03 120.0 4.70 0.81 19.40 6.83 18.50 5.73

2.25 463.1 2.97 2.42 3.96 17.40 1.57 6.66 544.0 15.83 100.1 2.83 192.4 11.87 3.82 95.7 17.52 31.42 27.80

2.86 170.8 3.05 27.98 0.20 4.25 1.42 6.08 175.8 7.68 29.85 3.92 56.45 9.47 0.74 24.98 12.60 9.24 35.22

3.02 143.1 5.36 24.87 0.38 27.55 1.41 9.27 159.0 7.61 24.63 5.19 141.2 15.48 1.63 121.0 8.14 32.67 26.33

3.13 384.4 5.36 24.96 9.37 62.53 3.27 8.71 441.1 33.02 222.6 4.22 51.38 27.82 4.43 45.08 27.70 74.16 67.13

3.25 191.1 2.64 20.55 0.26 7.58 1.33 4.45 200.0 11.51 31.07 4.95 165.6 6.18 0.98 104.3 3.51 50.59 27.89

1.20 178.0 4.85 24.90 0.66 12.60 0.63 1.28 62.40 7.20 17.00 5.08 82.9 2.82 0.69 25.90 3.09 4.71 10.40

1.20 36.28 1.43 7.82 0.06 13.25 0.36 0.80 30.20 3.81 6.77 1.55 72.35 0.98 0.38 11.30 5.25 9.78 3.04

1.29 331.7 1.75 3.89 0.08 11.47 0.59 3.34 359.4 4.27 22.34 1.70 166.0 3.15 1.00 13.84 1.76 8.26 8.99

1.47 101.3 4.84 15.00 0.33 9.25 0.89 5.05 111.2 40.00 13.38 1.59 99.0 4.14 1.39 26.25 3.21 17.25 18.25

(or increasing f values). Using the geochemical classification diagram of Herron (1988) the Kudankulam sandstones are classified as arkose, subarkose, and litharenite, except for two samples that fall in the sublitharenite field (Fig. 4). This classification is generally consistent with the petrographic data because on a QFL diagram these samples fall in the subarkose and sublitharenite fields (Pettijohn et al. 1972). Slight enrichment of SiO2 (wt. %) content in litharenite (the mean with one-standard-deviation value being 56 6 4; number of samples n 5 9) and arkose (55 6 7, n 5 20), as compared to subarkose (51 6 8, n 5 14) and sublitharenite (48.2 6 1.2, n 5 2) can be attributed to the variation of quartz in these sandstones. The average Na2O content for the Kudankulam sandstones (arkose 0.84 6 0.33, n 5 20; subarkose 0.78 6 0.34, n 5 14; litharenite 0.73 6 0.36, n 5 9; and sublitharenite 0.66 6 0.16, n 5 2) is less than 1%. The depletion of Na2O (, 1%) in all groups of sandstones (Table 1) can be attributed to a relatively smaller amount of Na-rich plagioclase in them, consistent with the petrographic data. K2O and Na2O contents and their ratios (Table 1) also are consistent with the petrographic observations, according to which K-feldspar dominates over plagioclase feldspar. Al2O3 content is high in arkose (6.4 6 1.5, n 5 20) and litharenite (5.9 6 0.8, n 5 9), but decreases in subarkose (3.5 6 1.1, n 5 14) and sublitharenite (2.7 6 0.6, n 5 2). Similarly, generally low concentrations of Fe2O3* and TiO2 in all Kudankulam sandstones reflect low abundances of heavy minerals such as Ti-bearing biotite, ilmenite, titanite, and titaniferous magnetite in the analyzed samples. Average K2O/ Al2O3 ratios in arkose, subarkose, and sublitharenite are greater than 0.3 except in litharenite (; 0.2), indicating that most K2O is present in Kfeldspar. On a K2O/Na2O—SiO2 /Al2O3 plot (Fig. 5) arkose and litharenite from the Kudankulam Formation are generally similar to the average arkose and litharenite rocks from the Bradore Formation in Labrador, Canada (Pettijohn et al. 1972). The Kudankulam sandstone samples fall away from average values of graywacke (Pettijohn et al. 1972), andesite, basalt 1 komatiite (Condie 1993), as well as average UCC (Taylor and McLennan 1985). These sandstones are only slightly higher in K2O/Na2O and SiO2 / Al2O3 ratios than gneisses, charnockites, and granites of the adjacent source area (Proterozoic Kerala Khondalite Belt of southern India; Chacko et al. 1992; Braun et al. 1996). Considering the somewhat mobile nature of these major elements, the above observations suggest that these igneous rocks could be a source for the Kudankulam sandstones. Trace Elements Trace-element concentrations of Kudankulam sandstones are reported in Table 2. In comparison with average upper continental crust (UCC) the

concentrations of most trace elements are generally low. The average relative concentration ratios lie between 0.1 and 1, except for Ni, with consistently much higher average relative concentration values (; 5.2–8.7), and low values of Rb (; 0.05 for litharenite), Zr (; 0.06–0.07 for subarkose and sublitharenite), and Hf (; 0.07 for litharenite), in some rock types (Fig. 6). In all groups of sandstones, Zr and Hf are somewhat depleted as compared to the other elements, particularly Pb, Y, and U (Fig. 6). Rare Earth Elements The results of REE analysis are given in Table 3 and are shown as chondrite-normalized patterns in Figure 7A for arkose and Figure 7B for other rock types. SREE concentrations vary widely in Kudankulam sandstones (SREE ; 15–148). This wide variation is also characteristic of individual rock types, e.g., arkose (; 15–148) and subarkose (; 20–108). All analyzed sandstone samples have SREE abundances less than the average UCC (; 143; Taylor and McLennan 1985) except one arkose sample (C3) with SREE 5 ; 148 (Table 3). All groups of the Kudankulam sandstones show slight LREE-enriched and relatively flat HREE patterns with negative Eu anomaly, except two subarkose samples (B2 and C7) with practically no Eu anomaly (Table 3; Fig. 7B). These two samples also contain higher contents of Sr, considering that both ions (Eu 21 and Sr 21 ) have comparable ionic sizes and thus tend to preferably substitute for Ca 21 in plagioclase (Gao and Wedepohl 1995). DISCUSSION

Tectonic Setting Roser and Korsch (1986) established a discrimination diagram using log(K2O/Na2O) versus SiO2 to determine the tectonic setting of terrigenous sedimentary rocks. These authors used CaO and LOI-free 100% adjusted data to determine their field boundaries (see Fig. 12b in Roser and Korsch 1986). Both parameters (SiO2 and log(K2O/Na2O) values) increase from volcanic-arc to active-continental-margin to passive-margin settings (Fig. 8A). Because all Kudankulam sandstone samples have a considerable amount of CaO (; 13.1–31.6, 19 6 5, n 5 45), the major-element data were recalculated to 100% and CaO and LOI-free basis before plotting them in Figure 8A. This diagram (Fig. 8A) shows a passive-margin setting for all groups of Kudankulam sandstones. Discrimination of tectonic settings on the basis of major-element data also was proposed by Bhatia (1983); it includes oceanic island arc, continental island arc, active continental margin, and passive margin. Most of the Kudankulam sandstone samples fall in the general area of passive-

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA

291

TABLE 2.—Extended. Subarkose

Litharenite

Sublitharenite

H6

F7

C5

F3

C7

G4

G5

1.57 106.5 1.83 10.41 0.08 6.15 0.68 1.54 118.0 4.67 10.50 1.95 69.82 1.23 0.51 18.46 12.43 9.70 4.97

1.68 286.0 1.43 8.21 0.28 7.43 0.35 0.75 15.20 9.35 21.70 2.01 110.0 1.08 0.82 16.70 4.79 23.70 19.60

1.89 120.0 1.45 22.70 0.06 9.34 1.37 2.07 96.3 7.03 16.90 1.53 96.3 3.02 1.47 32.60 11.50 9.48 5.01

2.74 156.0 3.61 26.70 0.10 26.40 1.56 3.51 28.50 10.00 25.60 3.05 61.10 4.70 1.07 22.00 15.20 10.30 25.40

3.02 140.0 2.01 6.98 0.24 4.81 1.03 9.73 110.0 3.08 33.70 1.12 170.0 1.52 0.53 45.70 18.60 9.07 28.00

2.74 98.5 5.03 16.00 0.36 8.65 0.43 10.50 108.0 4.25 6.09 1.48 123.0 1.15 0.94 20.10 6.02 20.70 19.30

2.02 111.0 3.15 28.83 0.10 14.65 0.64 2.34 118.0 9.92 17.06 2.21 100.4 2.63 1.65 30.75 9.47 27.39 11.77

margin and active-continental-margin fields of the TiO2 versus Fe2O3* 1 MgO plot (Fig. 8B), but mostly in the passive-margin field of the Al2O3 / SiO2 versus Fe2O3* 1 MgO diagram (Fig. 8C). Low values of Al2O3 /SiO2 ratio are an indication of the quartz enrichment in the Kudankulam sandstones (Fig. 8C). These discrimination diagrams suggest that the tectonic setting of the Cauvery Basin in which Kudankulam sandstones were deposited was that of a passive margin, which is consistent with general geology of the southern Indian Peninsula. Weathering in the Source Area Petrographic data show that K-feldspar dominates over plagioclase, which may result from intense weathering in the source area or from diagenetic alteration. The latter can be ruled out by the presence of abundant carbonate cement that developed probably during early diagenesis (Armstrong Altrin Sam et al. 2001). The intensity and duration of weathering in sedimentary rocks can be evaluated by examining the relationships among alkali and alkaline earth elements (Nesbitt and Young 1982, 1996). Feldspars are by far the most abundant of the reactive minerals. Consequently, the dominant process during chemical weathering of the upper crust is the alteration of feldspars and the neoformation of clay minerals. During weathering, calcium, sodium, and potassium are largely removed from feldspars (Nesbitt et al. 1980). The amount of these elements surviving in the soil profiles and in the associated sediments is a quantitative index of the intensity of weathering (Fedo et al. 1996; Nesbitt et al. 1997). A good measure of the degree of chemical weathering can be obtained by calculation of the chemical index of alteration (CIA; Nesbitt and Young 1982) using the formula (molecular proportions) CIA 5 [Al2 O3 /(Al2 O3 1 CaO* 1 Na2 O 1 K2 O)] 3 100 where CaO* is the amount of CaO incorporated in the silicate fraction of the rock. CIA values for the Kudankulam sandstones vary from ; 34.9 to 50.4 (44.5 6 3.9, n 5 45; Table 1). For each group of Kudankulam sandstones CIA varies as follows: arkose (45.9 6 3.4, n 5 20); subarkose (41.7 6 3.5, n 5 14); litharenite (46.7 6 2.6, n 5 9); and sublitharenite (40.6 6 4.6, n 5 2). The CIA values for the Kudankulam sandstones also are plotted in Al2O3-(CaO* 1 Na2O)-K2O (A–CN–K) compositional space (molecular proportions) in Figure 9. The compositions of average gneisses, charnockites, and granites from the source area (Chacko et al. 1992; Braun

FIG. 6.—Multi-element normalized diagram for the Kudankulam sandstones, normalized against average upper continental crust (Taylor and McLennan 1985), using the following values (in ppm): Co 5 10, Ni 5 20, Cr 5 35, V 5 60, Sr 5 350, Rb 5 112, Ba 5 550, Pb 5 20, Zr 5 190, Y 5 22, Nb 5 25, Hf 5 5.8, Th 5 10.7, and U 5 2.8. Two horizontal lines for rock/upper continental crust values of 1 and 0.1 are included for reference.

et al. 1996) and UCC (Taylor and McLennan 1985) are shown, also for comparison. In the A–CN–K diagram all the Kudankulam sandstones plot close to the plagioclase K-feldspar line, as well as to the source rocks, suggesting a low degree of chemical weathering of the Kudankulam sandstones. It is also indicated by their low values of the plagioclase index of alteration (PIA; Fedo et al. 1995) (41 6 7, n 5 45; Table 1), calculated by the following equation (molecular proportions): PIA 5 [(Al2O32K2O)/ (Al2O3 1 CaO* 1 Na2O2K2O)] 3 100, and are consistent with the CIA values. However, the CIA values (mean ; 44.5) of the Kudankulam sandstones are still slightly lower than those (close to 50) of upper-continental-crust and Proterozoic rocks (Fig. 9). Thus, the low CIA values of the Kudankulam sandstones do not reflect the general chemical weathering conditions in the source region, which can be inferred from the petrographic observations. This is probably due to the sedimentary sorting effect. Physical sorting of sediment during transport and deposition leads to concentration of quartz and feldspar with some heavy minerals in the coarse fraction and of secondary lighter and more weatherable minerals in the suspended-load sediments (Nath et al. 2000; Singh and Rajamani 2001b; Gu et al. 2002). Furthermore, there is a significant positive correlation (Fig. 10A) between the CIA and the mean grain size MZ (expressed in f) for the Kudankulam sandstones. The linear correlation coefficient (r 5 0.72, n 5 45) is statistically significant at a very strict significance level of 0.001 (or confidence level of 99.9%). For individual rock types, this correlation is significant only for arkose (r 5 0.82, n 5 20) but not for other rock types (subarkose, r 5 0.40, n 5 14; litharenite, r 5 0.39, n 5 9). The increase in CIA with decreasing particle diameter (expressed in millimeters) for the arkoses of the Kudankulam sandstone samples could suggest that the intensity of weathering increases from medium to fine or very fine sand. Th/U in sedimentary rocks is of interest because weathering and recycling is expected to result in oxidation and removal of U with a resultant increase in this ratio. Although highly reduced sedimentary environments can have enriched U leading to low Th/U ratios, weathering tends to result in oxidation of insoluble U41 to soluble U61 with loss of solution and elevation of Th/U ratios (McLennan et al. 1990; McLennan and Taylor 1980, 1991). The Th/U ratios in the Kudankulam sandstones range from 1.22 to 12.80 (Table 3), with an overall mean value of 4.3 6 3.0 (n 5 21). Upper crustal igneous rocks have Th/U averaging about 3.8, with considerable scatter (Taylor and McLennan 1985; Condie 1993; McLennan 2001). Considering the average Th/U ratio of the Kudankulam sandstones

292

J.S. ARMSTRONG-ALTRIN ET AL. TABLE 3.—Rare-earth-element concentrations in ppm for sandstones of the Kudankulam Formation. Arkose

Rock Type Sample # MZ La Ce Pr Nd Sm Eu Gd Tb Dy Er Tm Yb Lu SREE (LREE/HREE)§ Eu/Eu* Cr/Th Cr/Ni Cr/V Y/Ni Th/Sc Th/U Th/Co La/Y La/Co La/Sc (La/Lu)cn† (Gd/Yb)cn†

Subarkose

E4

A2

D4

D2

C2

H8

D1

E6

H10

C3

E2

F6

B2

B4

1.02 3.84 5.98 0.80 2.85 0.51 0.08 0.60 0.90 0.53 0.29 0.04 0.24 0.03 15.42 7.40 0.44 6.92 0.22 0.76 0.19 0.47 4.47 0.64 0.76 2.91 2.11 13.29 2.03

1.26 11.21 9.63 0.98 4.51 0.91 0.07 0.68 0.08 0.42 0.43 0.06 0.26 0.03 28.93 13.35 0.26 5.82 0.57 1.41 0.37 1.44 1.69 1.36 1.27 6.44 6.84 38.79 2.12

1.38 10.60 12.00 1.24 6.24 1.08 0.11 0.83 0.19 0.98 0.45 0.05 0.31 0.03 33.33 10.25 0.34 4.18 0.20 0.72 0.07 1.24 3.93 4.84 1.50 10.19 2.60 36.68 2.16

1.64 6.93 21.20 1.07 4.21 0.57 0.11 0.92 0.12 0.82 0.32 0.04 0.28 0.05 35.96 12.63 0.46 2.78 0.44 1.08 0.21 2.68 8.66 1.75 0.99 2.30 3.52 14.39 2.66

2.00 12.00 18.20 1.63 8.76 1.40 0.35 2.02 0.30 1.42 1.02 0.15 0.88 0.10 47.15 6.74 0.64 2.19 0.34 0.53 0.22 2.32 5.80 1.78 1.76 4.55 5.91 12.46 1.86

2.25 15.83 32.75 3.26 11.40 2.49 0.29 2.50 0.42 2.66 1.75 0.27 1.92 0.30 73.69 6.36 0.35 0.20 0.004 0.03 0.03 4.19 3.11 4.00 0.90 5.33 5.59 5.48 1.06

2.86 22.85 44.96 5.06 18.36 3.12 0.55 2.85 0.41 2.26 1.15 0.15 0.96 0.13 100.91 11.41 0.55 2.96 0.16 1.12 0.07 2.42 12.80 3.11 1.81 7.49 5.83 18.25 2.41

3.02 25.36 52.79 6.22 22.19 3.65 0.21 2.80 0.36 1.79 0.88 0.11 0.78 0.11 115.70 15.59 0.19 1.61 0.16 0.21 0.05 2.98 9.50 2.89 3.12 4.73 4.89 23.93 2.91

3.13 27.70 56.37 5.89 19.96 4.21 0.40 4.25 0.72 4.51 2.93 0.46 3.29 0.50 127.56 6.51 0.29 0.90 0.06 0.55 0.06 6.59 6.28 5.19 1.00 5.17 6.56 5.75 1.05

3.25 37.07 53.46 7.68 29.60 5.54 1.19 6.26 0.88 5.03 2.72 0.34 2.04 0.29 147.99 7.22 0.62 3.33 0.10 0.20 0.02 1.25 6.31 2.34 10.56 14.04 7.49 13.27 2.49

1.20 3.97 13.10 1.97 4.32 0.89 0.16 1.46 0.23 1.38 0.80 0.13 0.70 0.08 28.09 4.79 0.43 8.83 0.40 0.96 0.05 0.56 4.09 0.58 0.78 0.82 0.78 5.15 1.69

1.20 4.96 7.23 1.02 3.74 0.69 0.12 0.81 0.12 0.71 0.43 0.06 0.36 0.05 19.70 6.66 0.49 7.98 0.26 0.69 0.17 0.63 2.58 0.69 3.20 3.47 3.20 10.30 1.82

1.29 12.39 20.89 2.22 7.59 1.15 0.34 0.78 0.09 0.44 0.17 0.02 0.12 0.02 45.81 26.49 1.04 1.23 0.01 0.28 0.005 1.85 3.15 1.80 7.29 7.08 7.29 64.31 5.27

1.47 9.06 16.63 2.04 7.33 1.22 0.11 1.01 0.13 0.66 0.30 0.04 0.26 0.03 38.22 14.57 0.29 3.62 0.13 0.57 0.03 2.60 2.98 0.86 5.70 1.87 5.70 31.35 3.15

§ (LREE/HREE) 5 S(La-Sm)/S(Gd-Lu);

† Subscript cn refers to chondrite-normalized values.

close to the upper-continental-crust values, it is likely that these sandstones were derived from the least weathered source rocks. Hydraulic Sorting It is widely accepted that hydraulic sorting can lead to variation in REE concentrations in sediments with different grain-size fractions and mineral contents (Cullers et al. 1975; Cullers et al. 1979; Gromet et al. 1984; McLennan 1989). The observed variations in the SREE content for arkose (73 6 47, n 5 10), subarkose (50 6 26, n 5 9), litharenite (51.3), and sublitharenite (56.7) could be either due to weathering or to variation in lithology. During weathering the REEs are relatively immobile, so only minor enrichment or loss is expected. However, the LREE and HREE show different types of behavior and may become fractionated (Cullers 1988; Cullers et al. 1997; Condie et al. 1995; Condie et al. 2001). Thus, we interpret the observed variations in SREE content of the Kudankulam sandstones to be due to variations in grain-size fractions of these sandstones. Another possibility for the variations in SREE content among the Kudankulam sandstones can be related to an influence of a quartz dilution effect on abundance of heavy minerals and/or clay. However, we have observed petrographically low abundance of clay and heavy minerals in the Kudankulam sandstones. So the differences in SREE content may be due to the variations in grain-size fractions (MZ ; 1.02–3.25 f) among the samples, which may cause an enrichment or depletion of SREE content depending on the actual grain-size values. This interpretation is supported by significant correlation between SREE versus MZ (Fig. 10B). The linear correlation coefficient (r 5 0.92, n 5 21) is statistically significant at the very strict significance level of 0.001 (or confidence level of 99.9%) for all the Kudankulam sandstones, and also for the individual rock types, e.g., arkose (r 5 0.98, n 5 10) and subarkose (r 5 0.94, n 5 9). The average SREE contents in fine grain-size fractions are about 3 to 4 times higher than those for medium grain-size fractions (Table 3). This strongly suggests that the REEs are hosted mainly in fine and very fine grain-size fractions than medium grain-size fractions as proposed by Cullers et al. (1979), Cullers et al. (1988), Condie (1991), and Mongelli et al. (1996).

Provenance The high-field-strength elements (HFSE) such as Zr, Nb, Hf, Y, Th, and U are preferentially partitioned into melts during crystallization (Feng and Kerrich 1990), and as a result these elements are enriched in felsic rather than mafic sources. Additionally, they are thought to reflect provenance compositions as a consequence of their generally immobile behavior (Taylor and McLennan 1985). The slightly higher contents of Nb, Y, U, and Th in the samples with higher SREE probably reflects a control by grainsize fractionation during transport, and may also suggest a contribution from a felsic source with high concentration of these elements. The depletion of Zr and Hf in all groups of sandstones could be related to the size variation (r 5 0.70 for Zr vs. mean grain size Mz, expressed in f units, and r 5 0.67 for Hf vs. Mz; n 5 21) and depletion of heavy-mineral fractions such as zircon in the Kudankulam sandstones. REE, Th, and Sc are quite useful for inferring crustal compositions, because their distribution is not significantly affected by diagenesis and metamorphism and is less affected by heavy-mineral fractionation than that for elements such as Zr, Hf, and Sn (Cullers et al. 1979; Bhatia and Crook 1986; Wronkiewicz and Condie 1987; Cox et al. 1995; McLennan 2001; Mongelli and Dinelli 2001). REE and Th abundances are higher in felsic than in mafic igneous source rocks and in their weathered products, whereas Co, Sc, and Cr are more concentrated in mafic than in felsic igneous rocks and in their weathered products. Furthermore, ratios such as Eu/Eu*, (La/ Lu)cn, La/Sc, Th/Sc, La/Co, Th/Co, and Cr/Th are significantly different in mafic and felsic source rocks and can therefore provide information about the provenance of sedimentary rocks (Cullers et al. 1988; Wronkiewicz and Condie 1989; Condie and Wronkiewicz 1990; Cullers 1994). In our study, the Eu/Eu*, (La/Lu)cn, La/Sc, Th/Sc, La/Co, Th/Co, and Cr/Th values of the Kudankulam sandstones are similar to the values for sediments derived from felsic source rocks than those for mafic source rocks (Table 4), suggesting that these sandstones probably were derived from felsic source rocks. Furthermore, the relative REE patterns and the size of the Eu anomaly also have been used to infer sources of sedimentary rocks (Taylor and

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA

293

TABLE 3.—Extended. Subarkose

Litharenite

Sublitharenite

H6

F7

C5

F3

C7

G4

G5

1.57 11.40 15.15 2.52 9.77 2.00 0.47 2.26 0.34 2.05 1.13 0.15 0.93 0.13 46.61 5.56 0.67 8.46 0.09 0.56 0.11 0.63 2.41 0.67 5.85 6.23 5.85 9.10 1.97

1.68 9.27 20.40 2.14 8.20 1.48 0.12 0.66 0.10 0.65 0.26 0.03 0.16 0.03 42.94 20.96 0.32 7.60 0.54 0.49 0.32 0.54 1.32 0.76 4.61 6.48 4.61 32.08 3.34

1.89 13.50 18.40 2.78 10.10 2.01 0.12 0.75 0.15 1.21 0.31 0.05 0.27 0.03 48.69 15.65 0.25 7.52 0.24 0.70 0.12 1.97 2.05 2.08 8.82 9.31 8.82 46.72 2.25

2.74 20.10 25.36 3.71 15.70 2.86 0.32 2.02 0.32 2.78 1.06 0.15 1.01 0.09 73.07 8.68 0.39 5.68 0.94 1.21 0.53 1.54 4.39 1.30 6.59 5.57 6.59 23.19 1.62

3.05 21.70 40.50 7.02 25.30 4.97 1.08 1.79 0.35 2.98 2.37 0.31 1.97 0.27 108.49 9.13 0.91 4.59 0.06 0.15 0.17 1.36 2.87 0.76 19.38 10.80 19.38 8.34 0.74

2.74 8.40 23.50 2.67 9.20 2.13 0.30 2.63 0.34 1.68 0.92 0.13 0.73 0.11 51.34 6.73 0.39 13.91 0.15 0.80 0.06 0.78 1.22 0.23 5.68 1.67 5.68 7.93 2.94

2.02 13.21 23.34 2.89 10.65 1.93 0.42 1.97 0.28 1.63 0.86 0.11 0.70 0.09 56.71 8.82 0.65 10.96 0.24 0.94 0.08 1.19 1.59 0.83 5.98 4.19 5.98 15.24 2.28

McLennan 1985; Wronkiewicz and Condie 1989). Felsic igneous rocks usually contain higher LREE/HREE ratios and negative Eu anomalies, and mafic igneous rocks contain lower LREE/HREE ratios with little or no Eu anomalies (Cullers 1994, 2000). Some tonalites or granodiorites derived from eclogite melting may contain very large LREE/HREE ratios with little or no Eu anomalies (Cullers and Graf 1984). In our study of the Kudankulam sandstones, their high LREE/HREE ratio (11 6 6, n 5 21) and a significant negative Eu anomaly (0.48 6 0.22, n 5 21) support felsic igneous rocks as a possible source (Table 3). This interpretation is in agreement with the result of a study on terrigenous sediments of Kudankulam limestones (Armstrong-Altrin et al. 2003). Source Rocks In Figure 11, we used the REE data for the Kudankulam sandstones compared with those for gneisses, charnockites, and granites of the southern Indian Proterozoic Kerala Khondalite Belt as a tool for determining their source rocks (Fig. 1A; Chacko et al. 1992; Braun et al. 1996) as well as with the UCC (Taylor and McLennan 1985). The general shapes of all REE patterns for the sandstones (Fig. 11A) are similar to the source rocks (Fig. 11B). However, in detail the sandstone samples have somewhat steeper patterns than the Kerala Khondalite Belt rocks. Clear negative Eu anomalies are present in most rock types, except for a granite sample. This suggests that the Kudankulam sandstones could have been derived by the contributions from nearby gneisses, charnockites, and granites (Kerala Khondalite Belt source area in Fig. 1A). Many samples have (Gd/Yb)cn ratios more than 2 (Table 3), suggesting that these sediments were derived from sources having somewhat depleted heavy rare earth elements whereas others have (Gd/Yb)cn ratios less than 2, which suggests that they were derived from less HREE-depleted Archean or post-Archean sources, or a combination of both. The average ratios of Proterozoic gneisses, charnockites, and granites from the source area (Chacko et al. 1992; Braun et al. 1996) also are shown in this plot. These rock types have (Gd/Yb)cn and Eu/Eu* ratios overlapping with the Kudankulam

FIG. 7.—Chondrite-normalized rare earth element plots for the Kudankulam sandstones with sample numbers and grain-size values (MZ 5 mean grain size, expressed in f units; Table 1) given next to the sample numbers; Chondrite normalization values are from Taylor and McLennan (1985). A) Arkose rocks; B) other rock types.

sandstones (Fig. 12), suggesting that the Kerala Khondalite Belt rocks (Fig. 1A) could have been the source rocks for the sandstones. On the basis of the geochemistry of immature flood-plain sediments of the Cauvery River, Singh and Rajamani (2001a, 2001b) interpreted the exposed rocks in southern India to have been uplifted and exposed because of the stress buildup by the Himalayan Orogeny. However, they stated that the time of this uplift might have been anytime from the Cenozoic to as late as the Quaternary. The present study can be used to constrain that the Proterozoic rocks in southern India were already exposed in late Miocene time and supplied sediments to the Kudankulam area. Ni Enrichment The ferromagnesian (or so-called compatible) trace elements Cr, Ni, Co, and V show generally similar behavior in magmatic processes, but they may be fractionated during weathering (Feng and Kerrich 1990). In the studied samples, Cr, Co, and V are slightly depleted and Ni is highly enriched with respect to the average composition of the UCC. This enrichment in Ni may suggest some input of mafic materials from the source terrane; however, the simultaneous depletion of Cr (16.08 6 8.56, n 5 21; Table 2), MgO (0.9 6 0.3, n 5 45; Table 1), and Cr/Th (5.3 6 3.6, n 5 21; Table 3) ratio suggests that other factors could have played a role in concentrating Ni in the sandstones. High concentrations of Ni (50–130 ppm) and Cr (112–225 ppm) in floodplain sediments of the Cauvery River (Fig. 1A), were reported and inter-

294

J.S. ARMSTRONG-ALTRIN ET AL.

FIG. 9.—CIA ternary diagram, Al2O3 (A)—CaO* 1 Na2O (CN)—K2O (K), after Nesbitt and Young (1982) (CaO* 5 CaO in silicate phase) showing 1Kudankulam sandstones—this study, as well as average compositions of different rock types: 2UCC (upper continental crust) from Taylor and McLennan (1985); 3Chacko et al. (1992); 4Braun et al. (1996).

FIG. 8.—Tectonic-setting discrimination diagrams for the Kudankulam sandstones. The tectonic settings are named in each plot. A) SiO2—(K2O/Na2O) (after Roser and Korsch 1986); B) Fe2O3* 1 MgO—TiO2 (after Bhatia 1983); C) Fe2O3* 1 MgO—Al2O3 /SiO2 (after Bhatia 1983).

preted to be a result of chemical weathering of mafic source rocks (Singh and Rajamani 2001a). However, this interpretation of mafic source rocks is not supported by concentrations of REE and other trace elements in the Kudankulam sandstones (Table 4). Garver et al. (1996) suggested that elevated Cr and Ni abundances (Cr .150 ppm and Ni .100 ppm) and low Cr/Ni ratios (between 1.3 and 1.5) are indicative of ultramafic rocks in the source area of shales, although Cr/ Ni ratios were much higher (. 3.0) for sandstones from the same area.

FIG. 10.—Bivariate plots for the Kudankulam sandstones. A) MZ—CIA [MZ 5 grain size expressed in f units; CIA (chemical index of alteration) 5 [Al2O3 /(Al2O3 1 CaO* 1 Na2O 1 K2O)] 3 100]. B) MZ—(REE)T (total REE content).

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA

295

TABLE 4.—Range of elemental ratios of Kudankulam sandstones in this study compared to the ratios in similar fractions derived from felsic rocks, mafic rocks, and upper continental crust. Elemental Ratio Eu/Eu* (La/Lu)cn La/Sc Th/Sc La/Co Th/Co Cr/Th 1 2 3

Range of Sandstones from Kudankulam Formation1 (n 5 21)

Range of Sediment from Felsic Sources 2

Range of Sediment from Mafic Sources 2

Upper Continental Crust3

0.19–1.04 5.15–64 0.78–19.4 0.47–6.59 0.82–14.0 0.23–5.19 0.20–13.9

0.40–0.94 3.00–27.0 2.50–16.3 0.84–20.5 1.80–13.8 0.67–19.4 4.00–15.0

0.71–0.95 1.10–7.00 0.43–0.86 0.05–0.22 0.14–0.38 0.04–1.40 25–500

0.63 9.73 2.21 0.79 1.76 0.63 7.76

This study. Cullers (1994, 2000); Cullers and Podkovyrov (2000); Cullers et al. (1988). McLennan (2001); Taylor and McLennan (1985).

Abundances of chromium and nickel in the Kudankulam sandstones are on average 16 ppm and 138 ppm, respectively; Cr/Ni ratios range from ; 0.004 to 0.936 (0.25 6 0.22, n 5 21; Table 3), but are mostly below 0.5. Although high Cr and Ni abundances are clearly suggestive of mafic and/ or ultramafic provenance, ultramafic lithologies in ophiolite sequences can attain Cr/Ni ratios of 10 or greater (e.g., Jaques et al. 1983) and accordingly the low Cr/Ni ratios in Kudankulam sandstones are not especially indicative

FIG. 12.—Plot of Eu/Eu* versus (Gd/Yb)cn for the Kudankulam sandstones. Fields are after McLennan and Taylor (1991). The average upper continental crust (UCC) and gneisses, charnockites, and granites from the Proterozoic Kerala Khondalite Belt of southern India also are included. 1Kudankulam sandstones, this study; 2Taylor and McLennan (1985); 3Chacko et al. (1992); 4Braun et al. (1996).

FIG. 11.—Chondrite-normalized REE patterns for: A) sandstone samples from this study and, for comparison, average upper continental crust (UCC) (1Kudankulam sandstones, this study; 2Taylor and McLennan 1985); B) gneisses, charnockites, and granites from the Proterozoic Kerala Khondalite Belt of southern India (3Chacko et al. 1992; 4Braun et al. 1996), which are used here to represent compositions of source rocks.

of an ultramafic provenance. Therefore, the hypothesis proposed by Garver et al. (1996) may not be applicable to the Kudankulam sandstones. The low Cr/V ratios ; 0.03 to 1.41 (0.66 6 0.37, n 5 21) and the variable Y/Ni ratios ; 0.005 to 0.533 (0.14 6 0.13, n 5 21; Table 3) indicate that no ophiolite component is present in the Kudankulam sandstones (e.g., Hiscott 1984). Relatively high Ni contents in the Kudankulam sandstones can be carried by orthopyroxene grains derived from charnockites. These orthopyroxene grains may eventually break down to pyrite, a mineral that may influence the ferromagnesian budget of sedimentary rocks. In fact, pyrite is a common authigenic phase that may scavenge these elements from pore waters after the breakdown of relatively unstable mafic minerals such as pyroxene and hornblende (Bock et al. 1998). These authors also showed that pyrite in their study had undetectable contents of Cr (, 170 ppm, the detection limit of electron microprobe) and low V (, 13 ; 50 ppm) abundances but generally higher Ni abundances (up to 880 ppm). Thus, the enriched Ni contents and associated low contents of Cr and V may suggest the presence of pyrite in Kudankulam sandstone samples. This should be confirmed in future by microprobe or X-ray analysis. Yet another possibility, which we consider to be the most likely explanation for the Ni positive anomaly in the Kudankulam sandstones (Fig. 6), is fractionation of garnet between the source rocks and these sandstones. Garnet is known to be present in the source rocks (Braun et al. 1996; Braun 1999) as well as in coastal sediments from southern India (Ramasamy et al. 1996; Sabeen et al. 2002). Garnet has very high Cr/Ni ratios (; 115; e.g., Glaser et al. 1999) and the partition coefficients for Ni, Cr, V, and Co (e.g., Sisson and Bacon 1992; Torres-Alvarado et al. 2003) differ in such a way that this mineral can strongly fractionate Ni from the other elements, as is observed in the Kudankulam sandstones. Thus, it is possible that the unusual Ni enrichment, unaccompanied by a similar enrichment in Co, Cr, and V, is due to the fractionation of garnet from the source rocks during transportation. Fractionation of garnets would also explain the small differences observed in REE patterns (Figs. 11A, B) between the sandstone samples and source rocks. Similarly, the slight shift of the Kudankulam sandstones towards higher (Gd/Yb)cn values as compared to the source rocks (Fig. 12) can also be readily explained by such a fractionation of garnet. This fractionation may be related to one or more of several factors, such as hydraulic sorting, mechanical abrasion, diminution of garnet grain size and isolation from the sand fraction, and chemical weathering (e.g.,

296

J.S. ARMSTRONG-ALTRIN ET AL.

Morton et al. 1994; Nesbitt et al. 1997; Chigira and Oyama 1999; Morton and Hallsworth 1999). Further research is needed to determine which of these processes was the most important for the garnet fractionation in the Kudankulam sandstones. CONCLUSIONS

The K2O and Na2O values and their ratios in Kudankulam sandstones indicate that K-feldspar dominates over plagioclase feldspar. The tectonicsetting discrimination diagrams support a passive-margin setting for the Kudankulam sandstones. The low CIA values (; 44.5) of the Kudankulam sandstones do not reflect the general conditions of chemical weathering in the source region. This probably is due to the sedimentary sorting effect. Physical sorting of sediment during transport and deposition leads to concentration of quartz and feldspar with some heavy minerals in the coarse fraction and of secondary lighter and more weatherable minerals in the suspended-load sediments. The increase in CIA with decreasing grain-size fractions reveals that the intensity of weathering increases from medium to fine or very fine sand. Eu/Eu*, (La/Lu)cn, La/Sc, Th/Sc, La/Co, Th/Co, and Cr/Th ratios and the REE patterns indicate the derivation of these sandstones from felsic igneous source rocks. The differences of the REE characteristics of the Kudankulam sandstones probably do not reflect changes in source composition; the differences can be explained by the variations in grain-size fractions among these sandstones, suggesting that the REE are hosted mainly in the fine and very fine size fractions than in the medium size fractions. We conclude that the source rocks included gneisses, charnockites, and granites. The present study suggests that the Proterozoic rocks in southern India were already exposed in late Miocene time and supplied sediments to the Kudankulam area. The Ni enrichment unaccompanied by an increase in Cr, Co, and V in the Kudankulam sandstones calls for new studies to understand this rare observation. ACKNOWLEDGMENTS

The authors are thankful to Prof. S.P. Mohan, Head, Department of Geology, University of Madras, for providing certain laboratory facilities through SAP-II and UGC COSIST programs. The first author wishes to express his gratefulness to Robert L. Cullers, P.K. Banerjee, J. Madhavaraju, and S. Srinivasalu for their useful suggestions and guidance during the course of this study. We are grateful to the reviewers Giovanni Mongelli and Salvatore Critelli and Editor Mary J. Kraus and Associate Editor Mark Johnsson for numerous helpful comments to improve our paper. Technical editing by John B. Southard is highly appreciated. This research was partly supported by Council of Scientific and Industrial Research (CSIR), New Delhi grant (24/239/98-EMR-II to SR) and Korea Science and Engineering Foundation (KOSEF) grant (2000-2-13100-003-5 to YIL). REFERENCES ARMSTRONG-ALTRIN, J.S., VERMA, S.P., MADHAVARAJU, J., LEE, Y.I., AND RAMASAMY, S., 2003, Geochemistry of upper Miocene Kudankulam limestones, southern India: International Geology Review, v. 45, p. 16–26. ARMSTRONG ALTRIN SAM, J., AND RAMASAMY, S., 1997, Petrography and major element geochemistry of bioclastic rocks around Kudankulam, Tamil Nadu: Indian Association of Sedimentologists, Journal, v. 16, p. 171–182. ARMSTRONG ALTRIN SAM, J., AND RAMASAMY, S., 1999, Granulometry, petrography, geochemistry, and depositional environments of sand-rich sequence of Kudankulam Formation, Tamil Nadu: Indian Association of Sedimentologists, Journal, v. 18, p. 187–200. ARMSTRONG ALTRIN SAM, J., RAMASAMY, S., AND MAKHNACH, A., 2001, Stable isotope geochemistry and evidence for meteoric diagenesis in Kudankulam Formation, Tamil Nadu: Geological Society of India, Journal, v. 57, p. 39–48. BHATIA, M.R., 1983, Plate tectonics and geochemical composition of sandstones: Journal of Geology, v. 91, p. 611–627. BHATIA, M.R., AND CROOK, A.W., 1986, Trace element characteristics of graywackes and tectonic setting discrimination of sedimentary basins: Contributions to Mineralogy and Petrology, v. 92, p. 181–193. BOCK, B., MCLENNAN, S.M., AND HANSON, G.N., 1998, Geochemistry and provenance of the Middle Ordovician Austin Glen Member (Normanskill Formation) and the Taconian Orogeny in New England: Sedimentology, v. 45, p. 635–655. BRAUN, I., 1999, Generation of leucogranites in the Kerala Khondalite Belt, southern India: Physics and Chemistry of the Earth (A), v. 24, p. 281–287.

BRAUN, I., RAITH, M., AND RAVINDRA KUMAR, G.R., 1996, Dehydration–melting phenomena in leptynitic gneisses and the generation of leucogranites: A case study from the Kerala Khondalite Belt, southern India: Journal of Petrology, v. 37, p. 1285–1305. BRUCKNER, H., 1988, Indicators for formerly higher sea levels along the east coast of India and on the Andaman Islands: Hamburger Geographische Studien, v. 44, p. 47–72. CHACKO, T., RAVINDRA KUMAR, G.R., MEEN, J.K., AND ROGERS, J.J.W., 1992, Geochemistry of high-grade supracrustal rocks from the Kerala Khondalite Belt and adjacent massif charnockites, south India: Precambrian Research, v. 55, p. 469–489. CHIGIRA, M., AND OYAMA, T., 1999, Mechanism and effect of chemical weathering of sedimentary rocks: Engineering Geology, v. 55, p. 3–14. CONDIE, K.C., 1991, Another look at rare earth elements in shales: Geochimica et Cosmochimica Acta, v. 55, p. 2527–2531. CONDIE, K.C., 1993, Chemical composition and evolution of upper continental crust: Contrasting results from surface samples and shales: Chemical Geology, v. 104, p. 1–37. CONDIE, K.C., AND WRONKIEWICZ, D.J., 1990, The Cr/Th ratio in Precambrian pelites from the Kaapvaal craton as an index of craton evolution: Earth and Planetary Science Letters, v. 97, p. 256–267. CONDIE, K.C., DENGATE, J., AND CULLERS, R.L., 1995, Behavior of rare earth elements in paleoweathering profile on granodiorite in the Front Range, Colorado, USA: Geochimica et Cosmochimica Acta, v. 59, p. 279–294. CONDIE, K.C., LEE, D., AND FARMER, G.L., 2001, Tectonic setting and provenance of the Neoproterozoic Uinta Mountain and Big Cottonwood groups, northern Utah: Constraints from geochemistry, Nd isotopes, and detrital modes: Sedimentary Geology, v. 141–142, p. 443– 464. COORAY, P.G., 1984, Geology with special reference to the Precambrian, in Fernando, C.H., ed., Ecology and Biogeography of Sri Lanka: Dr. W. Junk B.V. Publishers, The Hague, The Netherlands, p. 1–34. COX, R., AND LOWE, D.R., 1995, A conceptual review of regional-scale controls on the composition of clastic sediment and the co-evolution of continental blocks and their sedimentary cover: Journal of Sedimentary Research, v. 65, p. 1–21. COX, R., AND LOWE, D.R., 1996, Quantification of the effects of secondary matrix on the analysis of sandstone composition, and a petrographic–chemical technique for retrieving original framework grain modes of altered sandstones: Journal of Sedimentary Research, v. 66, p. 548–558. COX, R., LOWE, D.R., AND CULLERS, R.L., 1995, The influence of sediment recycling and basement composition on evolution of mudrock chemistry in the southwestern United States: Geochimica et Cosmochimica Acta, v. 59, p. 2919–2940. CULLERS, R.L., 1988, Mineralogical and chemical changes of soil and stream sediment formed by intense weathering of the Danberg granite, Georgia, USA: Lithos, v. 21, p. 301–314. CULLERS, R.L., 1994, The controls on the major and trace element variation of shales, siltstones, and sandstones of Pennsylvanian–Permian age from uplifted continental blocks in Colorado to platform sediment in Kansas, USA: Geochimica et Cosmochimica Acta, v. 58, p. 4955– 4972. CULLERS, R.L., 2000, The geochemistry of shales, siltstones and sandstones of Pennsylvanian– Permian age, Colorado, USA: Implications for provenance and metamorphic studies: Lithos, v. 51, p. 181–203. CULLERS, R.L., AND GRAF, J.L., 1984, Rare-earth elements in igneous rocks of the continental crust: intermediate and silicic rocks—ore petrogenesis, in Henderson, P., ed., Rare Earth Element Geochemistry: Amsterdam, Elsevier, p. 275–316. CULLERS, R.L., AND PODKOVYROV, V.N., 2000, Geochemistry of the Mesoproterozoic Lakhanda shales in southeastern Yakutia, Russia: Implications for mineralogical and provenance control, and recycling: Precambrian Research, v. 104, p. 77–93. CULLERS, R.L., BASU, A., AND SUTTNER, L.J., 1988, Geochemical signature of provenance in sand-size material in soils and stream sediments near the Tobacco Root batholith, Montana, USA: Chemical Geology, v. 70, p. 335–348. CULLERS, R.L., BOCK, B., AND GUIDOTTI, C., 1997, Elemental distributions and neodymium isotopic compositions of Silurian metasediments, western Maine, USA: Redistribution of the rare earth elements: Geochimica et Cosmochimica Acta, v. 61, p. 1847–1861. CULLERS, R.L., CHAUDHURI, S., ARNOLD, B., LEE, M., AND WOLF, C.W., 1975, Rare-earth distributions in clay minerals and clay-sized fractions of Lower Permian Havensville and Eskridge shales of Kansas and Oklahoma: Geochimica et Cosmochimica Acta, v. 39, p. 1691–1703. CULLERS, R.L., CHAUDHURI, S., KILBANE, N., AND KOCH, R., 1979, Rare earths in size fractions and sedimentary rocks of Pennsylvanian–Permian age from the mid-continent of the USA: Geochimica et Cosmochimica Acta, v. 43, p. 1285–1302. DICKINSON, W.R., 1970, Interpreting detrital modes of graywacke and arkose: Journal of Sedimentary Petrology, v. 40, p. 695–707. DICKINSON, W.R., AND SUCZEK, C.A., 1979, Plate tectonics and sandstone compositions: American Association of Petroleum Geologists, Bulletin, v. 63, p. 2164–2182. FEDO, C.M., NESBITT, H.W., AND YOUNG, G.M., 1995, Unraveling the effects of potassium metasomatism in sedimentary rocks and paleosols, with implications for weathering conditions and provenance: Geology, v. 23, p. 921–924. FEDO, C.M., ERIKSSON, K.A., AND KROGSTAD, E.J., 1996, Geochemistry of shales from the Archaean (; 3.0 Ga) Buhwa Greenstone Belt, Zimbabwe: Implications for provenance and source-area weathering: Geochimica et Cosmochimica Acta, v. 60, p. 1751–1763. FENG, R., AND KERRICH, R., 1990, Geochemistry of fine-grained clastic sediments in the Archean Abitibi greenstone belt, Canada: Implications for provenance and tectonic setting: Geochimica et Cosmochimica Acta, v. 54, p. 1061–1081. FOLK, R.L., 1966, A review of grain-size parameters: Sedimentology, v. 6, p. 73–96. FOLK, R.L., AND WARD, W.C., 1957, Brazos river bar, a study in the significance of grain-size parameters: Journal of Sedimentary Petrology, v. 27, p. 3–26. FRALICK, P.W., AND KRONBERG, B.I., 1997, Geochemical discrimination of clastic sedimentary rock sources: Sedimentary Geology, v. 113, p. 111–124.

GEOCHEMISTRY OF SANDSTONES FROM SOUTHERN INDIA GAO, S., AND WEDEPOHL, K.H., 1995, The negative Eu anomaly in Archean sedimentary rocks: implication for decomposition, age and importance of their granitic sources: Earth and Planetary Science Letters, v. 133, p. 81–94. GARVER, J.I., ROYCE, P.R., AND SMICK, T.A., 1996, Chromium and nickel in shale of the Taconic Foreland: A case study for the provenance of fine-grained sediments with an ultramafic source: Journal of Sedimentary Research, v. 66, p. 100–106. GAZZI, P., 1966, Le arenarie del flysch sopracretaceo dell’Appennino modensese: Correlazioni con il flysch di Monghidoro: Mineralogica et Petrographica Acta, v. 12, p. 69–97. GLASER, S.M., FOLEY, S.F., AND GU¨NTHER D., 1999, Trace element compositions of minerals in garnet and spinel peridotite xenoliths from the Vitim volcanic field, Transbaikalia, eastern Siberia: Lithos, v. 48, p. 263–285. GOODWIN, A.M., 1991, Precambrian Geology; The Dynamic Evolution of the Continental Crust: London, Academic Press, 666 p. GROMET, L.P., DYMEK, R.F., HASKIN, L.A., AND KOROTEV, R.L., 1984, The North American shale composite: Its compilation, major and trace element characteristics: Geochimica et Cosmochimica Acta, v. 48, p. 2469–2482. GU, X.X., LIU, J.M., ZHENG, M.H., TANG, J.X., AND QI, L., 2002, Provenance and tectonic setting of the Proterozoic turbidites in Hunan, south China: Geochemical evidence: Journal of Sedimentary Research, v. 72, p. 393–407. HERRON, M.M., 1988, Geochemical classification of terrigenous sands and shales from core or log data: Journal of Sedimentary Petrology, v. 58, p. 820–829. HISCOTT, R.N., 1984, Ophiolitic source rocks for Taconic-age flysch: Trace-element evidence: Geological Society of America, Bulletin, v. 95, p. 1261–1267. HOLLAND, H.D., 1978, The Chemistry of the Atmosphere and Oceans: New York, Wiley, 351 p. JAQUES, A.L., CHAPPELL, B.W., AND TAYLOR, S.R., 1983, Geochemistry of cumulus peridotites and gabbros from the Marum Ophiolite Complex, northern Papua New Guinea: Contributions to Mineralogy and Petrology, v. 82, p. 154–164. JARVIS, K.E., 1988, Inductively coupled plasma mass spectrometry: A new technique for the rapid or ultra level determination of the rare-earth elements in geological materials: Chemical Geology, v. 68, p. 31–39. MCLENNAN, S.M., 1989, Rare earth elements in sedimentary rocks: Influence of provenance and sedimentary processes: Reviews in Mineralogy, v. 21, p. 169–200. MCLENNAN, S.M., 2001, Relationships between the trace element composition of sedimentary rocks and upper continental crust: Geochemistry Geophysics Geosystems, v. 2, paper number 2000GC000109 [8994 words, 10 figures, 5 tables]. Published April 20, 2001. MCLENNAN, S.M., AND TAYLOR, S.R., 1980, Rare earth element–thorium correlations in sedimentary rocks and the composition of the continental crust: Geochimica et Cosmochimica Acta, v. 44, p. 1833–1839. MCLENNAN, S.M., AND TAYLOR, S.R., 1991, Sedimentary rocks and crustal evolution: tectonic setting and secular trends: Journal of Geology, v. 99, p. 1–21. MCLENNAN, S.M., TAYLOR, S.R., MCCULLOCH, M.T., AND MAYNARD, J.B., 1990, Geochemical and Nd–Sr isotopic composition of deep-sea turbidites: Crustal evolution and plate tectonic associations: Geochimica et Cosmochimica Acta, v. 54, p. 2015–2050. MIDDLETON, G.V., 1960, Chemical composition of sandstones: Geological Society of America, Bulletin, v. 71, p. 1011–1026. MONGELLI, G., AND DINELLI, E., 2001, The geochemistry of shales from the ‘‘Frido Unit,’’ Liguride Complex, Lucanian Apennines, Italy: Implications for provenance and tectonic setting: Ofioliti, v. 26, p. 457–466. MONGELLI, G., CULLERS, R.L., AND MUELHEISEN, S., 1996, Geochemistry of Late Cretaceous– Oligocene shales from the Varicolori Formation, southern Apennines, Italy: Implications for mineralogical, grain-size control, and provenance: European Journal of Mineralogy, v. 8, p. 733–754. MORTON, A.C., AND HALLSWORTH, C.R., 1999, Processes controlling the composition of heavy mineral assemblages in sandstones: Sedimentary Geology, v. 124, p. 3–29. MORTON, A.C., HUMPHREYS, B., DHARMAYANTI, D.A., AND SUNDORO, 1994, Palaeogeographic implications of the heavy mineral distribution in Miocene sandstones of the North Sumatra Basin: Journal of Southeast Asian Earth Sciences, v. 10, p. 177–190. NATH, B.N., KUNZENDORF, H., AND PLU¨GER, W.L., 2000, Influence of provenance, weathering, and sedimentary processes on the elemental ratios of the fine-grained fraction of the bedload

297

sediments from the Vembanad lake and the adjoining continental shelf, southwest coast of India: Journal of Sedimentary Research, v. 70, p. 1081–1094. NESBITT, H.W., AND YOUNG, G.M., 1982, Early Proterozoic climates and plate motions inferred from major element chemistry of lutites: Nature, v. 299, p. 715–717. NESBITT, H.W., AND YOUNG, G.M., 1996, Petrogenesis of sediments in the absence of chemical weathering: effects of abrasion and sorting on bulk composition and mineralogy: Sedimentology, v. 43, p. 341–358. NESBITT, H.W., FEDO, C.M., AND YOUNG, G.M., 1997, Quartz and feldspar stability, steady and non-steady-state weathering, and petrogenesis of siliciclastic sands and muds: Journal of Geology, v. 105, p. 173–191. NESBITT, H.W., MARKOVICS, G., AND PRICE, R.C., 1980, Chemical processes affecting alkalies and alkaline earths during continental weathering: Geochimica et Cosmochimica Acta, v. 44, p. 1659–1666. PARAMASIVAM, T.S., AND SRINIVASAN, R., 1980, Sedimentary limestones of Tisaiyanvilai area, Tirunelveli district, Tamil Nadu: Oil and Natural Gas Commission of India, Bulletin, v. 17, p. 183–186. PETTIJOHN, F.J., POTTER, P.E., AND SIEVER, R., 1972, Sand and Sandstones: New York, SpringerVerlag, 618 p. PIPER, D.Z., 1974, Rare earth elements in the sedimentary cycle: A summary: Chemical Geology, v. 14, p. 285–304. RAMANATHAN, S., 1979, Tertiary formations of south India: Geological Survey of India, Miscellaneous Publications, v. 45, p. 165–180. RAMASAMY, S., AND ARMSTRONG ALTRIN SAM, J., 1998, Inferences on rhodoids from Neogene carbonates of Kudankulam, Tamil Nadu, India: Geological Society of India, Journal, v. 52, p. 341–344. RAMASAMY, S., ARMSTRONG ALTRIN SAM, J., AND CHANDRASEKAR, K., 1996, Heavy mineral distribution around Manavalakurichi, Kanyakumari District, Tamil Nadu, India: Indian Association of Sedimentologists, Journal, v. 15, p. 29–42. ROSER, B.P., AND KORSCH, R.J., 1986, Determination of tectonic setting of sandstone–mudstone suites using SiO2 content and K2O/Na2O ratio: Journal of Geology, v. 94, p. 635–650. SABEEN, H.M., RAMANUJAM, N., AND MORTON, A.C., 2002, The provenance of garnet: constraints provided by studies of coastal sediments from southern India: Sedimentary Geology, v. 152, p. 279–287. SAHANI, A., 1979, Miocene vertebrates from the coastal Tertiary rocks of Peninsular India and Sri Lanka: Geological Survey of India, Miscellaneous Publications, v. 45, p. 197–205. SHAPIRO, L., 1975, Rapid analysis of silicate, carbonate, and phosphate rocks—revised edition: U.S. Geological Survey, Bulletin, 1401, p. 1–76. SHAPIRO, L., AND BRANNOCK, W.W., 1962, Rapid analysis of silicate, carbonate, and phosphate rocks: U.S. Geological Survey, Bulletin, 1144, p. 1–56. SINGH, P., AND RAJAMANI, V., 2001a, Geochemistry of the floodplain sediments of the Kaveri River, southern India: Journal of Sedimentary Research, v. 71, p. 50–60. SINGH, P., AND RAJAMANI, V., 2001b, REE geochemistry of recent clastic sediments from the Kaveri floodplains, southern India: Implication to source area weathering and sedimentary processes: Geochimica et Cosmochimica Acta, v. 65, p. 3093–3108. SISSON, T.W., AND BACON, C.R., 1992, Garnet/high-silica rhyolite trace element partition coefficients measured by ion microprobe: Geochimica et Cosmochimica Acta, v. 56, p. 2133– 2136. TAYLOR, S.R., AND MCLENNAN, S.M., 1985, The Continental Crust; Its Composition and Evolution: London, Blackwell, 312 p. TORRES-ALVARADO, I.S., VERMA, S.P., PALACIOS-BERRUETE, H., GUEVARA, M., AND GONZA´LEZ-CASTILLO, O.Y., 2003 DCpBASE: a database system to manage Nernst distribution coefficients and its application to partial melting modeling: Computers and Geosciences, v. 29, n. 9, p. 1191–1198. WRONKIEWICZ, D.J., AND CONDIE, K.C., 1987, Geochemistry of Archean shales from the Witwatersrand Supergroup, South Africa: Source-area weathering and provenance: Geochimica et Cosmochimica Acta, v. 51, p. 2401–2416. WRONKIEWICZ, D.J., AND CONDIE, K.C., 1989, Geochemistry and provenance of sediments from the Pongola Supergroup, South Africa: Evidence for a 3.0-Ga-old continental craton: Geochimica et Cosmochimica Acta, v. 53, p. 1537–1549. Received 8 October 2002; accepted 28 August 2003.