Geochemical tracers of source rocks in a Cretaceous to Quaternary sedimentary sequence (Eastern Sierras Pampeanas, Argentina)

Geochemical tracers of source rocks in a Cretaceous toQuaternary sedimentary sequence (Eastern Sierras Pampeanas,

Argentina)

E.L. Piovanoa,*, 1, G. Roma n Rossb, 1, S. Ribeiro Guevarab, 2, M.A. Arribe reb, 2,P.J. Depetrisa, 1

aFacultad de Ciencias Exactas, FõÂsicas y Naturales, Universidad Nacional de CoÂrdoba, Velez Sars®eld, 299, 5000 CoÂrdoba, ArgentinabLaboratorio de AnaÂlisis por ActivacioÂn NeutroÂnica RA-6, Centro AtoÂmico Bariloche, CNEA, 8400 San Carlos de Bariloche, Argentina

Abstract

Metamorphic rocks, granitic rocks, and sediments from the Eastern Sierras Pampanas, Argentina, were analyzed for majorand trace element concentrations, including rare earth elements (REE). Parental rocks exhibit distinctive REE normalized

diagram patterns and elemental ratios, and some elemental ratios reveal signi®cant di�erences between rock sources. Forexample, ratios such as Th/Sc, Cr/Th, and La/Cr have a mean value of 0.7, 8.4 and 0.4 in metamorphic rocks, whereas graniticrocks exhibit means of 1.4, 0.7 and 4.9, respectively. These ratios are also useful in linking detrital materials with the

corresponding parental rocks. Metamorphic sources yield sediments with lower Th/Sc and La/Cr, and higher Cr/Th ratios thansediments derived from granitic sources. REE and other elements are enriched in the silt-size fraction, whereas they are dilutedby quartz in the sand-size fraction.

The size of the Eu/Eu� anomaly can be used as a stratigraphical correlation tool in the sedimentary record: Cretaceous rocksshow a mean value of 0.920.1, whereas Tertiary rocks have a mean value of 2.920.3. The Eu anomaly in Quaternary andmodern sediments ranges from 0.5 to 0.8. # 1999 Elsevier Science Ltd. All rights reserved.

1. Introduction

A major goal in sedimentary studies is to ®nd help-

ful tools to determine environmental conditions (Folk

and Ward, 1957; Passega, 1964; Friedman, 1967), and

the provenance of sediments (Suttner, 1974; Argast

and Donnelly, 1987; Rooney and Basu, 1994; Cullers,

1995). Source rock types, physicochemical conditions

of depositional settings, and diagenetic processes may

have a controlling in¯uence on the composition of

clastic sedimentary rocks. In analyzing the sedimentary

material, geochemical tracers can be used to identify

source rocks and weathering processes (Cullers et al.,

1988; Nesbitt, 1979). In combination with traditional

methods, which apply facies analysis, geochemical

indexes provide an improved and more in-depth analy-

sis of source rock identi®cation. The best geochemical

indicators are those that are least a�ected by weather-

ing processes and represent the composition of the

source. Such ``immobile'' elements include the rare

earth elements (REE) patterns and ratios of La or Th

to Sc, Co or Cr (Taylor and McLennan, 1985; Condie

et al., 1995; Liu et al., 1993; Cullers, 1994a,b; 1995).

In the present work, we have investigated a terrige-

nous sedimentary sequence of Cretaceous to

Quaternary age derived from a known plutonic±meta-

morphic complex in order to establish geochemical

changes during sediment production. We have also

evaluated the relationship between chemical compo-

sition and texture to determine the controlling e�ect of

transport processes and grain-size.

Journal of South American Earth Sciences 12 (1999) 489±500

0895-9811/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.

PII: S0895-9811(99 )00031 -0

* Corresponding author.

E-mail address: [email protected] (E.L. Piovano).1 CONICET2 Instituto Balseiro

2. Geological setting

Located in the Eastern Sierras Pampeanas, CentralArgentina, the Sierras de Co rdoba basement mainlyconsists of a plutonic±metamorphic complex (Gordilloand Lencinas, 1979) of Lower Paleozoic age which wasintruded by granitoids during various times (Rapela etal., 1991, 1998a). The basement mainly includes poly-metamorphic rocks, granitoid plutons, and subordinatebasic and ultrabasic rocks (Baldo et al., 1996; Rapelaet al., 1998b) (Figs. 1 and 2).

In the study area, two ranges compose the EasternSierras Pampeanas: the Sierra Chica and the SierraGrande. The Sierra Chica Range, lying towards theeastern portion, is composed mainly of metaigneousand metasedimentary rocks. In the RõÂ o SuquõÂ a area(Gordillo and Lencinas, 1979; Baldo et al., 1996;Rapela et al., 1998b) the basement is composed ofCambrian ortho-gneisses, para-gneisses, schists, mig-matites, marbles and amphibolites, and Ordoviciantrondhjemites and tonalites.

West of Sierra Chica, the Achala batholith outcropsin Sierra Grande. It is a major multiphase granitic

complex of Devonian to Carboniferous age (Rapela etal., 1991) characterized by various petrological andgeochemical attributes (Lira and Kirschbaum, 1990,Rapela et al., 1991; Demange et al., 1996).

The sedimentary sequence studied (Fig. 2) is locatedon the eastern slope of the Sierra Chica. The sedimen-tary sequence begins with the red-bedded EarlyCretaceous Salda n Formation (250 m thick), whichdisconformably overlies the basement (Piovano, 1996).This unit consists of conglomerates, sandstones andmudstones, and it has been interpreted as having beendeposited in arid alluvial fan environments (Piovano,1995). The Salda n Formation deposition took place intwo sequences separated by an alkaline volcanic event,inferred by the presence of alkali basalt boulders in theuppermost sequence (Piovano, 1996).

The Villa Belgrano Formation (Tertiary?) crops outa few kilometers eastward and is composed of red-colored ®ne conglomerates, sandstones, and mudstonesaccumulated in alluvial fan and braided ¯uvial en-vironments. This formation has an observable thick-ness of 10 m.

Unconformably overlying Tertiary rocks, the clastic

Fig. 1. Geological map of the study area (after Gordillo and Lencinas, 1979), and sample locations. Insert at the lower right side shows an enlar-

gement of Quebrada del rõÂ o SuquõÂ a.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500490

Quaternary sequence begins with the EstanciaBelgrano Formation (Early Pleistocene; Santa Cruz,1972; Piovano et al., 1992). This Formation (8 mthick) contains stacked ®ning-upward cycles composedof ®ne gravel or sand at the base and mud at the top.These sediments have accumulated in distal areas ofalluvial fans. The Pampiano Formation (MiddlePleistocene), unconformably overlies the EstanciaBelgrano Formation, and is represented by unstrati®edclayey-silt or sandy-silt indicating eolian deposition. Ithas an observable thickness of 7 m. The inter®ngeringand overlying RõÂ o Primero Formation (Middle±LatePleistocene) has a measurable thickness of 3 m. Its tex-ture ranges from gravel to sand, and corresponds to abraided ¯uvial facies. The General Paz Formation(Late Pleistocene±Early Holocene) has a thickness of4 m and is mainly composed of silt, having isolatedgravel/sand lenses. These types of sediments are widelyknown as loess-like deposits. Eolian processes were notthe dominant mechanism in the Pampiano and

General Paz Formation. Occurrences of sand andgravel lenses within the ®ne sediments indicate rework-ing by shallow streams.

3. Samples and methodology

The stratigraphic sequence was sampled at the Rõ oSuquõ a Valley, Sierra Chica (Co rdoba, Argentina, Fig.1), where the sedimentary rocks are almost ¯at lying.Sampling took place at the following: (a) Plutonic±metamorphic complex (S1±S7); (b) Salda n Formation,considering the sequences prior to vulcanism (S9±S14),post volcanic (S15±S17) and the basaltic rocks (S8); (c)Villa Belgrano Formation (S18 and S19), and (d) theQuaternary Formations: Estancia Belgrano (S20),Pampiano (S21), Rõ o Primero (S22) and General Paz(S23). Present-day sediments were also sampled andare represented by bottom sediments taken from theMal Paso Reservoir (S24) and ¯uvial channel sedi-ments (S25) collected in the Sierra Grande.

The chemical composition of sediments was deter-mined on the less than 2-mm grain-size fraction toavoid the bias produced by rock fragments in coarsesamples. Grain-size determinations were carried out onthe chemically analyzed fraction, by dry sieving and bythe pipette methods (McManus, 1988). Mean grain-size, expressed in Phi units (Mz), sorting (S1), skewness(Sk1) and kurtosis (KG) were calculated according toFolk and Ward (1957). Samples were texturally classi-®ed following Shepard's scheme (1954).

Samples were ground in a tungsten carbide ball-milland elemental concentrations were determined byinstrumental neutron activation analysis (INAA) usingthe absolute parametric method in the RA-6 Barilocheresearch reactor. Roma n Ross et al. (1995) reporteddetails on the methodology and results obtained instandard reference materials.

4. Results and discussions

4.1. Source rock compositions

Lithofacies and paleocurrents in the Cretaceous toQuaternary sequence have revealed that the underlyingplutonic±metamorphic complex supplied terrigenousclastic materials (Piovano et al., 1992; Piovano, 1995).The most voluminous units in Sierra Chica are ortho-gneisses, para-gneisses, schists, migmatites, and theremaining units (i.e., marbles, amphibolites, andOrdovician granitoids) are volumetrically small.Devonian±Carboniferous granitoids are dominant inSierra Grande.

Geochemical features in metamorphic rocks weredetermined on a gneiss (S1), and a diatexite (S2), both

Fig. 2. Stratigraphic scheme for the study area. Collected samples

are listed for each Formation.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500 491

Table

1

Elementalconcentrationsandelem

entalratiosofthesamples.Metamorphic

rock

averagedoes

notincludemarble

composition

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500492

with sedimentary protoliths (Baldo et al., 1996; Rapelaet al., 1998b), and on an amphibolite (S3) and amarble (S4). Elemental concentrations and elementalratios are exhibited in Table 1. Three samples analyzedby Rapela et al. (1998b) were also considered: a diatex-ite (RSU-65), a gneiss (RSU-71) and an ortho-gneiss(RSU-77).

Excluding marble, which is di�erentiated by a lowconcentration of REE, the rest of the metamorphicrocks show similar patterns in the REE chondrite-nor-malized diagram (Fig. 3a), with light REE (LREE)concentrations greater than those of heavy REE(HREE). The La/Yb ratio in metasedimentary rocksranges from 3.01 to 11.27. Meta-igneous rocks show

large variations in REE patterns (Rapela et al., 1998b)with La/Yb ratios ranging from 5 to 15 in basic andintermediate members, to 55 in the more evolved mem-bers. Whether igneous or sediment derived, the meta-morphic rocks considered in this paper show acompositional similarity that allows them to bereferred to as a single population. In this way, a meta-morphic rock mean composition was calculated. Thechemical data for marble was not considered to calcu-late this mean.

The average granitic rock composition (Table 1) wascalculated using data from a monzogranite, a tonalite,and a pegmatite, all from the Sierra Chica. Thesamples used were considered to be representative on

a Total sum of REE in ppm.b Sedimentary environments are PAF: Proximal alluvial fan, MAF: Middle alluvial fan, DAF: Distal alluvial fan, E: Eolian and F: Fluvial.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500 493

the basis of previous petrographic knowledge. TheREE concentrations and the LREE/HREE ratio in in-dividual samples show marked ¯uctuations (e.g., La/Yb ranges from 4.03 to 36.99) and the Eu anomalyranges from almost zero to negative (Fig. 3b).

The Achala granitic complex presents a large com-

positional variation in major and trace elementsaccording to the di�erent existing facies (Morteani etal., 1995). Distribution patterns of REE concentrationsare highly variable with a La/Yb ratio ranging from 10to 56.

The granitic rocks average shows lower REE con-

Fig. 3. Chondrite-normalized REE patterns of source rocks and their derived sediments. Chondrite data from Evensen et al., (1978). (a)

Metamorphic rocks, (b) Granitic rocks and basalt, (c) Salda n Formation, (d) Tertiary rocks, (e) Quaternary sediments, (f) Present-day sediments.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500494

centrations than the metamorphic rocks average, andsome elemental ratios in both rocks are signi®cantlydi�erent (Table 1). For example, the F test shows thatCr/Th, La/Cr, and Sm/Nd ratios in metamorphic andgranitic rocks exhibit signi®cant di�erences between

their respective variances at the 95% con®dence level.Th/Sc, whose variances are not signi®cantly di�erent(95% con®dence level) in both rock groups, show sig-ni®cant di�erences (95% con®dence level) in their re-spective means (Student's t test). Metamorphic rocksexhibit lower Th/Sc and La/Cr ratios and higher Cr/Th values than igneous rocks. The La/Yb ratio doesnot conclusively separate granitic from metamorphicrocks. Marble is characterized by low elemental con-centrations with respect to the remaining source rocks.Hence, its control on the sedimentary composition iseasily masked by the in¯uence of metamorphic origneous rocks.

Cretaceous basalts (S8 in Table 1) are of the alkalinetype and exhibit the highest REE concentration in thesample set, thus di�erentiating volcanic rocks from theremaining source rocks (Fig. 3b). The REE patterns inthe normalized diagram are similar to patterns ofCretaceous basalts in the Pampean Range (Kay andRamos, 1996).

4.2. Chemical composition of sediments

Salda n Formation samples have similar REE nor-malized patterns (Fig. 3b), whether they belong to theupper post-volcanic or to the lower pre-volcanicsequences. Samples corresponding to the post-volcanicsedimentary deposition (S15±S17) do not show REEcompositional features or elemental ratios that mightsuggest a signi®cant contribution from volcanic rocks.For example, the Cr/Th ratio, and the SREE di�ersigni®cantly in basalts (ca 500 and 390, respectively)from sedimentary rocks (ca 6 and 170, respectively).

REE patterns of the Salda n Formation are similarto those of metamorphic rocks (Fig. 3). Their meanvalues for the Eu/Eu� and Ce/Ce� (0.9 and 1.0 respect-ively) are slightly higher than those from the plutonic±metamorphic complex. Like the REE normalized pat-terns, the multi-elemental diagram of average compo-sition of the Salda n Formation (normalized withrespect to the upper continental crust, UCC) is similarto that of the metamorphic rock average (Fig. 4a).Multi-elemental diagrams of individual samples alsoexhibit similar shapes (Fig. 4b) suggesting a uniformmetamorphic source during the deposition of theSalda n Formation.

With the exception of Sm and Tb values, REE nor-malized concentrations of the Tertiary Villa BelgranoFormation exhibit the lowest SREE concentration ofthe sedimentary sequence and depict REE patternswhich are closer to the granitic rocks compositionalarea (Fig. 3c). Although not consistent with a graniticsource, the rocks show markedly positive Eu anomalieswhich are higher than any other of the measuredsamples (Eu/Eu� values of 2.66 and 3.12 in S18 andS19 respectively). Elemental patterns (Fig. 4c) also

Fig. 4. UCC normalized extended diagrams. (a) Metamorphic and

granitic rocks, (b) Salda n Formation, (c) Tertiary rocks, (d)

Quaternary sediments, (e) Present-day sediments. Upper Continental

Crust (UCC) data from Taylor and McLennan (1985).

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500 495

show the lowest concentrations in most of themeasured elements, which are closer to a granitic thanto a metamorphic rock composition.

The mineralogy of Cretaceous and Tertiary sedi-ments can be related to the observed dissimilarity inSREE concentrations. Cretaceous samples consist, indecreasing order of abundance, of altered feldspar, det-rital quartz and biotite. Heavy mineral concentrationsare low and consist of garnet, amphiboles, pyroxenesand opaques. Tertiary rocks are composed of quartz,microcline, plagioclase and biotite. Accessory heavyminerals are rare and include tourmaline and rutile inquartz grains. Pyroxenes, amphiboles and opaqueshave not been recorded in the sample set. The abun-dance of quartz and feldespars and the depletion inma®c minerals, could explain the low REE concen-trations in the Tertiary sedimentary rocks. The REEand Th may be concentrated in monazite, sphene, alla-nite, apatite, zircon, or garnet, and large variations inthe abundance of heavy minerals result in large com-positional variation (Cullers et al., 1988).

The Eu anomaly value is directly linked to the pre-sence of Ca-plagioclase. The anomalies are the resultof fractionation which separates granitic melts fromresidues containing feldspar, mainly Ca-plagioclase,which is the main host of Eu2+ (Taylor andMcLennan, 1988). The low abundance of feldspar (8±10%) in Tertiary rocks is not consistent with theextreme positive Eu anomaly. As the enrichments ordepletions of REE and other soluble major elementscould be related to pH (Nesbitt, 1979), the Euanomalies in Tertiary rocks could be attributed topost-depositional processes. Although both formationsare red-colored due to the presence of hematite, thecolor of the Cretaceous rocks is more intense becauseit has more ferric pigment than the lighter-coloredTertiary Formation (note %Fe in Table 1). Thesedi�erences in color could be due to the leaching ofhematite, that could be produced in acidic environ-ments with low E values (Faure, 1992). The low Ferelative abundance is correlated with low REE concen-trations, and this is re¯ected by a positive correlationcoe�cient between these variables (SREE vs %Fe;r = 0.74, p < 0.05). Van der Wiejden and Van derWiejden (1995) noted that the mobility of REE duringweathering, under reducing conditions, produces a cur-ious trend in the REE patterns with selective variationsin the abundance of REE. They have also shown acase of depletion in Sm and Tb and the immobility ofEu in weathered materials. It is most likely that, asreported by Van der Wiejden and Van der Wiejden(1995), the examined Tertiary rocks probably under-went changes in the redox condition during post-depositional processes which produced the mobiliz-ation of iron together with some REE, especially Smand Tb, thus enhancing the Eu anomaly values.

Quaternary and recent sediments have REE normal-ized patterns (Fig. 3d) which are similar to other REEpatterns observed in terrigenous sedimentary rocks(e.g., Taylor and McLennan, 1988; Liu et al., 1993).Within Quaternary sediments, alluvial (EstanciaBelgrano Formation, S20), and the eolian sediments(Pampiano, S21 and General Paz Formations, S23)show similar REE pattern shapes, that can be relatedto the pattern of the modern-day sample S24 (Fig. 3e),which was derived from a mixed granitic±metamorphicenvironment. The sample series (from S20 to S25)

Fig. 5. Relationships among Th and Nd contents and depositional

environments and processes. Grain-size compositions of the samples

are presented in Table 1. (5a) Salda n Formation. (5b) Tertiary±

Quaternary deposits (including present-day).

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500496

exhibit variable multi-elemental diagrams (Fig. 4d±e).When such diagrams are used to relate sediments totheir sources, the conclusion is similar to that obtainedwhen REE normalized patterns are used.

Although eolian sediments (S21 and S23) havedi�erent stratigraphic ages, they contain similar REEnormalized patterns, hence suggesting a similar prove-nance. Even though these eolian deposits had an im-portant input of allochtonous materials includingvolcanic ash, their resulting REE patterns are close tothe metamorphic rock mean pattern. The reworkingaction of shallow streams on the eolian deposit, withsediments mainly derived from the Sierra Chica (domi-nantly metamorphic rocks), probably produced amasking e�ect which hid the allochtonous material sig-nature of the eolian deposits. Isolated gravel lenses inthe clay/silty deposits support this assumption.

Chemical composition of riverbed samples shows thepassage from dominantly granitic terrain (S25), tometamorphic dominance (S24). The shifting is revealedin the REE patterns (Fig. 3e), showing enrichment inthe concentration of REE as the relative signi®canceof metamorphic source increases. Sample S22 (RõÂ oPrimero Formation of Upper Pleistocene age), whichwas deposited by the same ¯uvial system, underwentthe longest transport over metamorphic terrain andshows the highest REE concentration of the set. Wehave observed that when metamorphic rocks are pre-sent they imprint their own chemical features on theresulting sediments.

4.3. Elemental concentrations, elemental ratios, and thetextural e�ect

Table 1 shows chemical compositions and texturaldata. Positive linear correlations exist between the silt-size percentage and most of the elemental concen-trations, indicating a chemical enrichment in this size-fraction. Cs and Rb, for instance, show positive corre-lation coe�cients with silt percentage (Cs vs %silt,r = 0.94; Rb vs %silt, r = 0.89; all p < 0.1).Accordingly, both elements are depleted in all samplesat higher sand percentages (Cs vs %sand, r=ÿ0.87;Rb vs %sand, r=ÿ0.79; all p < 0.1).

Th and Nd concentrations change signi®cantly withsilt abundance. As the percentage of silt varies due tothe competence of the transporting agent, the Salda nFormation samples are discriminated according totheir position within the paleoenvironment (Table 1).In contrast with sediments accumulated in distal zones(Fig. 5a), those sediments deposited in proximal set-tings (with lesser silt %) have lower Th and Nd values.Tertiary and Quaternary formations depict a similartrend (Fig. 5b).

A marked covariance of Hf and Zr (r= 0.94;p < 0.1) suggest an identical source. Hf and Zr posi-tive anomalies, as observed in our study, are usuallyattributed to the presence of zircon (Condie, 1993). Ybis also concentrated in zircons; large increases in Zrconcentrations are associated with smaller incrementsin Yb concentrations. Such covariance suggests that

Fig. 6. Th/Sc and La/Cr variability in sediments and source rocks. Sediments derived from metamorphic rocks show lower Th/Sc and La/Cr

ratios than those supplied by a granitic sources. MRA: Metamorphic rocks average, GRA: Granitic rocks average, SFA: Salda n Formation aver-

age.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500 497

Yb is associated with zircon, a member of the heavymineral suite which hosts HREE (Sholkovitz, 1990).

In agreement with metamorphic source rocks, allsediment samples exhibit high Ta concentrations.Sediments derived entirely from granitic areas (i.e.,S25) show Ta concentrations similar to their sources.

Positive correlation between SREE, and concen-trations of some REE (i.e., Ce, Nd, Eu and Tb) withrespect to silt percentages, suggest an enrichment ofsuch elements in this size fraction (e.g., SREE vs%silt, r = 0.42, p < 0.1). The relationship is attributedto the presence of hornblende in controlling light andmiddle REE in the silt-size fractions. Correlation coef-®cients between SREE or individual REE concen-trations and clay percentages lacked signi®cance. Lowconcentrations for most trace-elements in the coarserthan 62.5 mm size-fraction is attributed to the higherquartz to other minerals ratios (SREE vs %sand,r=ÿ0.35; p < 0.2). According to the textural classi®-cation (Table 1), sandstones and sands are the mostdepleted in REE. Skewness (Sk1) exhibits a signi®cantand positive correlation coe�cient with the SREE(r = 0.5; p < 0.1), thus indicating an increase ofSREE contents in the positively skewed or ®nersamples.

Elemental ratios (Table 1) vary according to di�er-ent sediment grain-sizes. The La/Sc ratio tends todecrease in ®ne grained sediments (La/Sc vs Mz,r=ÿ0.48; p < 0.1) due to the higher Sc concentrationin the ®ner classes (e.g., Sc vs Mz, r = 0.44; p < 0.1).The La/Yb ratio is also di�erent in di�erent grain-sizes. Yb concentrations increase with increasing siltrelative content (Yb vs %silt, r = 0.4; p < 0.1).Therefore the La/Yb ratio decreases in the ®nestsamples (La/Yb vs Mz, r=ÿ0.5; p < 0.1). Neither theStudent's t-test nor the F-test allows the use of La/Scand La/Yb ratios to characterize granitic or meta-morphic sources because their variances and means arenot signi®cantly di�erent.

As discussed above, Th/Sc, Cr/Th, La/Cr and Sm/Nd ratios are signi®cantly di�erent (95% con®dencelevel) in metamorphic and granitic rocks (Table 1).Although these ratios are not transferred unchangedfrom parent materials to sediments (see above, the in-¯uence of silt percentages in Th and Nd concentrationsand the relationship between Sc vs Mz), they allow thelinking of sediments to parental rocks. The Cr, La,and Sm concentrations do not show correlations withsedimentary size fractions.

Some average elemental ratios are useful to thecharacterization of parental source rocks and their det-rital products. Sediments derived from metamorphicsources, for example, can be identi®ed because theyshow lower values in the Th/Sc and La/Cr ratios, andhigher Cr/Th ratios than those produced by a graniticoutput. Also, Sm/Nd ratios in sediments tend to be

higher from metamorphic rocks sources than fromigneous sources.

A plot of Th/Sc vs La/Cr illustrates the two groupsof sources (Fig. 6). The derived sediments are in inter-mediate positions between source rock-types.According to these ratios, Cretaceous rocks could berelated to a dominant metamorphic source rock (Th/ScSalda n Fm=0.85 and La/CrSalda n Fm=0.67). Althoughthey were probably a�ected by leaching processes thatchanged the original concentrations, Tertiary rocks arealso linked to a metamorphic source (Th/ScTertiaryrocks=0.33 and La/CrTertiary rocks=0.67). BothCretaceous and Tertiary rocks show Cr/Th valuescomparable to metamorphic rocks.

The composition of sample S25 (a ``control sample''taken in a modern-day ¯uvial channel) is representa-tive of sediments derived from a dominantly graniticsource. A comparison of S25 to the remainingQuaternary sediments suggests that the latter were de-rived from dominant metamorphic sources (S21 andS24) and mixed metamorphic±granitic sources (S21,S22 and S23).

The Eu anomaly could not be linked with texturalparameters. Additionally, high Ce anomalies are corre-lated with high silt contents (r= 0.52; p < 0.1).

5. Conclusions

Trace-element concentrations in the studied sedimen-tary sequence provide a useful signature which can beused to identify provenance. Granitic rocks showlower REE concentrations than metamorphic rocks,consequently REE normalized and multi-elemental dia-grams reveal patterns that can be used in linking sedi-ments to their probable source rocks.

Statistical analyses reveal that Cr/Th, La/Cr, Th/Scand, to a lesser extent, Sm/Nd, exhibit signi®cantdi�erences in both groups of source rocks (i.e., meta-morphics and granitics). Moreover, such ratios appearto be a diagnostic tool which can be used to identifysources. For example, low La/Cr and Th/Sc and highCr/Th values are a signature of metamorphic prove-nance. Also, an enrichment in the concentration ofREE in ¯uvial sediments is indicative of an increase inthe relative signi®cance of the metamorphic source.The longest transport over metamorphic terrain pro-duces the highest REE concentration.

Sediment sources can best be discriminated by plot-ting Th/Sc and La/Cr ratios. For example Cretaceous,Tertiary rocks, and some Quaternary sediments, allwith dominant metamorphic source rock, group nearthe composition of metamorphic rocks. Sediments withmixed sources are separated into a di�erent group anda sample representative of sediments derived from a

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500498

dominantly granitic source is markedly di�erentiatedin the plot.

Positive correlations between silt percentages andSREE, or the concentrations of Ce, Nd, Eu and Tb,indicate an enrichment of all these elements in the silt-size fraction. Low concentrations of trace elements athigher sand percentages are attributed to the dilutingpresence of quartz. In spite of the fact that elementalratios are a�ected by the grain-size distribution of sedi-ments, they can be used as geochemical tracers.

The Eu anomaly could not be linked with texturalparameters. In contrast, Ce anomalies are a�ected bysilt contents. The Eu anomaly can be used in the stu-died sequence as a stratigraphical correlation tool dif-ferentiating Tertiary rocks (mean value of 2.920.3)from the remaining sediments.

Acknowledgements

We thank R. Cullers and C.W. Rapela for theircareful revision of an earlier version of this manu-script. We also thank A.J. Kestelman for his interestin this work. The helpful assistance of A. Kirschbaum,D. Gaiero, E. Martinez and B. Theisen is gratefullyacknowledged. We are especially grateful to the per-sonnel of the RA-6 for the irradiation of samples. Thiswork has been partially funded by Argentina'sCONICET (PIP 4829).

References

Argast, S., Donnelly, T.W., 1987. The chemical discrimination of

clastic sedimentary components. Journal of Sedimentary

Petrology 57, 813±823.

Baldo, E.G., Casquet, C., Galindo, C., 1996. El metamor®smo de la

Sierra Chica de Co rdoba (Sierras Pampeanas). Argentina.

Geogaceta 19, 51±54.

Condie, K.C., 1993. Chemical compositions and evolution of the

upper continental crust: contrasting results from surface samples

and shales. Chemical Geology 104, 1±37.

Condie, K.C., Dengate, J., Cullers, R.L., 1995. Behavior of rare

earth elements in a paleoweathering pro®le on granodiorite in the

Front Range, Colorado, USA. Geochimica et Cosmochimica

Acta 59, 279±294.

Cullers, R.L., 1994a. The control on the major and trace element

variation of shales, siltstones, and sandstones of Pennsylvanian±

Permian age from uplifted continental blocks in Colorado to plat-

form sediment in Kansas, USA. Geochimica et Cosmochimica

Acta 58, 4955±4972.

Cullers, R.L., 1994b. The chemical signature of source rocks in size

fractions of Holocene stream sediment derived from metamorphic

rocks in the Wet Mountains region, Colorado, U.S.A. Chemical

Geology 113, 327±343.

Cullers, R.L., 1995. The controls on the major-and-trace element

evolution of shales, siltstones and sandstones of Ordovician to

Tertiary age in the Wet Mountains region, Colorado, U.S.A.

Chemical Geology 123, 107±131.

Cullers, R.L., Basu, A., Suttner, J.L., 1988. Geochemical signature

of provenance in sand-size material in soil and stream sediments

near the Tobacco Root Batholith, Montana, U.S.A. Chemical

Geology 70, 335±348.

Demange, M., Alvarez, J.O., Lo pez, L., Zarco, J., 1996. Existencia

de series magma ticas diferentes en el batolito de Achala

(Co rdoba, Argentina). In: XII Congreso Geolo gico Argentino y

II Congreso de Exploracio n de Hidrocarburos. Actas, vol. IV,

pp. 23±29.

Evensen, N.M., Hamilton, P.J., O'Nions, R.K., 1978. Rare earth

abundance in chondritic meteorites. Geochimica et Cosmochimica

Acta 42, 1199±1212.

Faure, G., 1992. Principles and Applications of Inorganic

Geochemistry. Macmillan, New York.

Folk, R.L., Ward, W.C., 1957. Brazos River bar: a study in the sig-

ni®cance of grain size parameters. Journal of Sedimentary

Petrology 27, 3±26.

Friedman, G.M., 1967. By name processes and statistical parameters

compared for size frequency distribution of beach and river

sands. Journal of Sedimentary Petrology 67, 327±354.

Gordillo, C.E., Lencinas, A. 1979. Sierras Pampeanas de Co rdoba y

San Luis. In: Turner, J.C. (Ed.), Segundo Simposio de GeologõÂ a

Regional Argentina, Tomo I, Academia Nacional de Ciencias,

Co rdoba, Argentina, pp. 577±650.

Kay, S., Ramos, V.A., 1996. El magmatismo creta cico de las Sierras

de Co rdoba y sus implicancias tecto nicas. In: XIII Congreso

Geolo gico Argentino y III Congreso de Exploracio n de

Hidrocarburos, Actas, vol. III, pp. 453±464.

Lira, R., Kirschbaum, A.M., 1990. Geochemical evolution of gran-

ites from the Achala batholith of the Sierras Pampeanas.

Geological Society of America 241, 67±76 Special paper.

Liu, C., Masuda, A., Okada, A., Yabuki, S., Zhang, J., Fan, Z.,

1993. A geochemical study of loess and desert sand in northern

China: implications for continental crust weathering and compo-

sition. Chemical Geology 106, 359±374.

McManus, J., 1988. Grain size determination and interpretation. In:

Tucker, M. (Ed.), Techniques in Sedimentology. Blackwell

Scienti®c Publications, Oxford, pp. 63±85.

Morteani, G., Preinfalk, C., Spiegel, W., Bonalumi, A., 1995. The

Achala Granitic Complex and the pegmatites of the Sierras

Pampeanas (Northwest Argentina): a study of di�erentiation.

Economic Geology 90, 636±647.

Nesbitt, H.W., 1979. Mobility and fractionation of rare earth el-

ement during weathering of a granodiorite. Nature 279, 206±210.

Passega, R., 1964. Grain size representation by CM patterns as a

geological tool. Journal of Sedimentary Petrology 384, 840±847.

Piovano, E.L., 1995. Los para metros fõ sicos y quõÂmicos y su vincula-

cio n con el modelo depositacional de la Formacio n Salda n

(Creta cico temprano), Sierra Chica de Co rdoba, Repu blica

Argentina. Asociacio n Argentina de Sedimentologõ a, Revista 2,

57±73.

Piovano, E.L., 1996. Correlacio n de la Formacio n Salda n (Creta cico

temprano) con otras secuencias de las Sierras Pampeanas y de las

Cuencas Chacoparanense y de Parana . Revista de la Asociacio n

Geolo gica Argentina 51, 29±36.

Piovano, E.L., Manzur, A., Pasquini, A., 1992. Ana lisis paleoam-

biental de las formaciones a¯orantes al oeste de la Ciudad de

Co rdoba. IV Reunio n Argentina de Sedimentologõ a III, 105±111.

Rapela, C.W., Pankhurst, R.J., Kirschbaum, A.M., Baldo, E.G.,

1991. Facies intrusivas de edad carbonõÂ fera en el batolito de

Achala: evidencia de una anatexis regional en las Sierras

Pampeanas. In: VI Congreso Geolo gico Chileno, Actas, pp. 40±

43.

Rapela, C.W., Pankhurst, R.J., Casquet, C., Baldo, E., Saavedra, J.,

Galindo, C., 1998a. Early evolution of the Proto-Andean margin

of South America. Geology 26, 707±710.

Rapela, C.W., Pankhurst, R.J., Casquet, C., Baldo, E., Saavedra, J.,

Galindo, C., Fanning, C.M. 1998b. The Pampean Orogeny of the

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500 499

southern proto-Andes: Cambrian continental collision in the

Sierra de Co rdoba. In: Pankhurst RJ, Rapela CW (Eds.), The

Proto-Andean Margin of Gondwana. Geological Society of

London, Special Publications, 142, 181±217.

Roma n Ross, G., Ribeiro Guevara, S., Arribe re, M.A., 1995. Rare

earth geochemistry in sediments of the Upper Manso River

Basin, RõÂ o Negro, Argentina. Earth and Planetary Science Letters

133, 47±57.

Rooney, C.B., Basu, A., 1994. Provenance analysis of muddy sand-

stones. Journal of Sedimentary Research A64 (1), 2±7.

Santa Cruz, J.N., 1972. GeologõÂ a al este de la Sierra Chica

(Co rdoba), Valle del Rõ o Primero. Boletõ n de la Asociacio n

Geolo gica de Co rdoba I (34), 102±109.

Shepard, F.P., 1954. Nomenclature based on sand±silt±clay ratios.

Journal of Sedimentary Petrology 24, 151±158.

Sholkovitz, E.R., 1990. Rare earth elements in marine sediments and

chemical standards. Chemical Geology 88, 333±347.

Suttner, L.J. 1974. Sedimentary petrographic provinces: an evalu-

ation. In: Ross, C.A. (Ed.), Paleogeographic Provinces and

Provinciality, Soc. Econ. Paleont. Mineral., Spec. Publ. 21, 75±

84.

Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its

Composition and Evolution. Blackwell Scienti®c Publications,

Oxford.

Taylor, S.R., McLennan, S.M., 1988. The signi®cance of the rare

earth in geochemistry and cosmochemistry. In: Gscheider Jr.,

K.A., Eyring, L. (Eds.), Handbook on the Physics and Chemistry

of Rare Earth, vol. II. Elsevier Science Publishers, Amsterdam,

pp. 485±578.

Van der Weijden, C.H., Van der Weijden, R.D., 1995. Mobility of

major, minor and some redox-sensitive trace elements and rare-

earth elements during weathering of four granitoids in Central

Portugal. Chemical Geology 125, 1498±1515.

E.L. Piovano et al. / Journal of South American Earth Sciences 12 (1999) 489±500500