Significance of geochemical signatures on provenance in intracratonic rift basins: Examples from the Iberian plate
Published:January 01, 2007
- PDF LinkChapter PDF
M. Ochoa, M.E. Arribas, J. Arribas, R. Mas, 2007. "Significance of geochemical signatures on provenance in intracratonic rift basins: Examples from the Iberian plate", Sedimentary Provenance and Petrogenesis: Perspectives from Petrography and Geochemistry, José Arribas, Mark J. Johnsson, Salvatore Critelli
Download citation file:
Following the Variscan orogeny, the Iberian plate was affected by an extensional tectonic regime from Late Permian to Late Cretaceous time. In the central part of the plate, NW-SE–trending rift basins were created. Two rifting cycles can be identified during the extensional stage: (1) a Late Permian to Hettangian cycle, and (2) a latest Jurassic to Early Cretaceous cycle. During these cycles, thick clastic continental sequences were deposited in grabens and half grabens. In both cycles, sandstone petrofacies from periods of high tectonic activity reveal a main plutoniclastic (quartzofeldspathic) character due to the erosion of coarse-grained crystalline rocks from the Hesperian Massif, during Buntsand- stein (mean Qm72F25Lt3) sedimentation and during Barremian–early Albian times (mean Qm81F18Lt1). Geochemical data show that weathering was more intense during the second rifting phase (mean chemical index of alteration [CIA]: 80) due to more severe climate conditions (humid) than during the first rifting phase (mean CIA: 68) (arid climate).
Ratios between major and trace elements agree with a main provenance from passive-margins settings in terms of the felsic nature of the crust. However, anomalies in trace elements have been detected in some Lower Cretaceous samples, suggesting additional basic supplies from the north area of the basin. These anomalies consist of (1) low contents in Hf, Th, and U; (2) high contents in Sc, Co, and Zr; and (3) anomalous ratios in Th/Y, La/Tb, Ta/Y, and Ni/V. Basic supplies could be related to the alkaline volcanism during Norian-Hettangian and Aalenian-Bajocian times. Geochemical composition of rift deposits has been shown to be a useful and complementary tool to petrographic deduction in provenance, especially in intensely weathered sediments. However, diagenetic processes and hydrothermalism may affect the original detrital deposits, producing changes in geochemical composition that mislead provenance and weathering deductions.
Studies in provenance have been mainly performed according to classic sandstone petrography. At present, several models are in common usage to deduce provenance parameters (source rock lithology, climate, weathering, transport, geotectonic setting) from petrographic analysis on sandstone framework (e.g., Basu et al., 1975; Dickinson and Suczek, 1979; Dickinson, 1985).
During the last two decades, the use of geochemical data for provenance inferences has experienced an important development (see McLennan et al., 1993), and models have been elaborated to decipher aspects concerning source lithology (Floyd and Leveridge, 1987; Gu et al., 2002), weathering (e.g., Nesbitt and Young, 1982; Taylor and McLennan, 1985), maturation during transport (Bhatia, 1983; Gu et al., 2002; Whitmore et al., 2004), and geotectonic setting (Bhatia and Taylor, 1981; Maynard et al., 1982; Bhatia, 1983, 1984; Roser and Korsch, 1985, 1986, 1988; Bhatia and Crook, 1986; McLennan and Taylor, 1991; Gu et al., 2002). Geochemical procedures generate quick and objective data, and they can be used on whole rock of clastic deposits from a wide range of grain sizes (Eynatten et al., 2003). In spite of these advantages, geochemical analysis may indicate a consistent loss of information about textures. Provenance origin of clasts (intrabasinal or extrabasinal, Zuffa, 1980; coeval or noncoeval, Zuffa, 1991) is indecipherable by whole-rock chemical analysis. In addition, diagenetic products are mixed with detrital material, and this fact may produce biased inferences on the provenance of original clastic material (García et al., 2004).
The Iberian Range is a linear structure trending NW-SE in the northeast edge of the Iberian microplate (Fig. 1); it is an intracratonic, folded segment of the Alpine Chain that developed as a rift basin (Iberian Basin) in two phases (e.g., Salas et al., 2001): the first phase was generated from the Early Permian to the Late Triassic, and the second phase of rifting occurred from Late Jurassic to early Albian time (Fig. 2). During these cycles, thick clastic sequences, from alluvial to lacustrine at the top, were deposited in grabens and half grabens. Both cycles evolved to periods of postrift thermal subsidence, where predominant shallow-marine carbonate sedimentation took place. During the Paleogene and Lower–Middle Miocene, compressive events caused structural inversion, folding, and thrusting.
Stratigraphy, sedimentology, and petrography of sediments generated during the two active rifting phases (Arribas, 1984; Arribas et al., 2003; Benito et al., 2001; Martín-Closas and Alonso-Millán, 1998; Mas et al., 2003; Ochoa et al., 2004) and the tectonic evolution of the basin (Guimerà et al., 1995; Salas et al., 2001; Guimerà et al., 2004) have been consistently analyzed. Thus, these deposits represent an excellent opportunity to contrast geochemical data analysis with valuable background information.
The principal aim of this paper is to evaluate the informative power of geochemical data from petrographically well-known examples of clastic sediments generated at different times, but in a similar geotectonic scenario: the intracratonic Iberian Rift. Furthermore, the contrast of geochemical signatures between sediments generated during the two active stages of rifting may contribute to a better understanding of the evolution of the Iberian Rift. Finally, the data obtained in this paper will increase general knowledge of intracratonic rift basins and will be applicable to future models of such basin types.
The study area is located in the northwest sector of the Iberian Range in central Spain, and it includes clastic deposits from the first (Permian to Triassic) and second (Late Jurassic to Early Cretaceous) rifting stages. These deposits outcrop in two different areas: the Moncayo area, which is composed of Permian to Triassic deposits, and Cameros Basin, where Upper Jurassic to Lower Cretaceous deposits appear (Fig. 1).
During the Late Permian–Triassic extensional stage (rift 1, Fig. 2), the reactivation of the wrench faults as normal faults induced the propagation of rift systems in the Iberian plate as the Iberian Trough (Sopeña and Sánchez-Moya, 1997; Mas et al., 2003). This extensional stage corresponds to the beginning of the Alpine sedimentary cycle and is represented by the clastic Saxonian and Buntsandstein facies. These facies are continental clastic sediments that infilled asymmetrical half grabens (Fig. 3A). Buntsandstein deposits are mainly continental red beds that form the base of a sequence that evolve into siliciclastic and carbonate tidal sediments (Muschelkalk facies). The clastic infill in the Moncayo area consists mainly of arkosic deposits arranged into five main lithostratigraphic units (Arribas, 1984, 1985). The thickness of these deposits varies from 100 m to more than 900 m due to differential subsidence of troughs.
The second stage of rifting (Late Jurassic to Early Cretaceous) was related to the opening of the Central Atlantic (rift 2, Fig. 2). As a consequence, the Cameros Basin was formed as an extensional rift basin above a south-dipping ramp on a blind, low-angle normal fault several kilometers deep in the basement (Alonso and Mas, 1993; Guimerà et al., 1995; Salas et al., 2001). According to Alonso and Mas (1993) and Mas et al. (2003), the sedimentary record (Tithonian–early Albian) constitutes a large megasequence bounded by two main unconformities at the base and at the top, and it can be further subdivided into eight depositional sequences separated by minor unconformities (DS-1 to DS-8 in Fig. 2). This study is focused on the maximum synrift filling stage (DS-4 to DS-7 [late Berriasian to early Aptian]; Fig. 3B) related to maximum tectonic activity of this rifting phase. The infill of the basin varies drastically in thickness, from nearly 100 m in the marginal areas of the basin (toward NE and SW) to 2200 m in the depocentral areas (central sector) (Fig. 3B). This record is constituted by fluvial sequences that consist of coarse deposits of conglomerates and channelized fluvial (mainly braided) sandstone bodies in proximal areas evolving to meandering and lacustrine facies in distal areas (Mas et al., 2003; Ochoa et al., 2004). During the Middle to Late Cretaceous, a low-grade metamorphic event (hydrothermalism) took place in depocentral areas and affected the sedimentary record (Casquet et al., 1992; Barrenechea et al., 1995; Alonso-Azcárate et al., 1999).
A total of 53 sandstone and shale samples was collected for geochemical analysis from several stratigraphic sections from both Permian-Triassic deposits in the Moncayo area (24 samples) and from Lower Cretaceous deposits (DS-4 to DS-7) in the Cameros Basin (29 samples). Sample locations are shown in Figure 3. Major and trace elements were determined for all samples. Analyses were performed at the Actlabs Laboratories (Canada) by Code 4Lithoresearch. All geochemical data are reported in Tables 1, 2 202, and 3 302.
For petrographic analysis of Permian-Triassic sandstones, one of the authors (J. Arribas) examined several databases from previous works (Arribas, 1984, 1987). This author provided thin sections of sandstones for analysis. For this paper, a new point counting method was performed on these samples following Gazzi-Dickinson criteria to obtain geotectonic inferences (Dickinson, 1985). In addition, Lower Cretaceous sandstones were analyzed petrographically following the same procedures and point-counting methods as those used for Permian-Triassic sandstones (Arribas et al., 2003; Ochoa et al., 2004).
Permian and Triassic Sandstones
Framework composition of sandstones varies from quartzose (mean Qm97F0Lt3; Qm—monocrystalline quartz; F—feldspars; Lt—total lithics) at the base of the succession (Saxonian facies; PS in Figs. 3A and 4A) to quartzofeldspathic (mean Qm72F25Lt3) petrofacies at the top of the sedimentary sequence (Buntsandstein facies; B-1 and B-2 in Figs. 3A and 4A).
Quartzose Saxonian petrofacies (Fig. 5A) are very mature texturally; they show evidence of maturation during transport (very well-sorted sediments and high values of quartz grain roundness) and recycling of metasediments in the Variscan basement (e.g., presence of inherited quartz overgrowth). Lithic rock fragments are scarce and consist mainly of low-grade metamorphic fragments (shales and chert); some quartzose sandstone fragments also occur. Syntaxial quartz overgrowth is the main interstitial cement in the sandstones. Framework composition of Saxonian sandstones is very homogeneous in all the Moncayo area. These sandstone petrofacies represent the initial stage of rift 1, when metasedimentary cover was eroded.
Sandstone composition of Buntsandstein facies represents a drastic change from the provenance of underlying sediments. Quartzofeldspathic petrofacies suggest that the contribution of coarse crystalline rocks from the Hesperian Massif (Fig. 1A) diluted the supplies from metasedimentary rocks (Arribas et al., 1985). The content of feldspar varies, showing a consistent increase of K-feldspar to the top (Figs. 4A and 5B). The presence of these K-feldspar–rich petrofacies may suggest both arid conditions of Buntsandstein sedimentation (Arribas, 1984) and nearness to source area. Syntaxial K-feldspar and quartz overgrowths and minor carbonate are the main interstitial cements. Matrix is constituted by illite and minor kaolinite minerals and is mainly of diagenetic origin: epimatrix (alteration of K-feldspars), pseudo-matrix (lithic rock fragment disaggregration), kaolinite pore-filling and illite pore-lining (Arribas, 1987).
The sandstone framework composition of depositional sequences 4, 5, 6, and 7 is very quartzose, with variable amounts of K-feldspar and lithics (Fig. 4B). In proximal areas (sections 1, 2, and 3 from CBA in Figs. 1 and 3), sandstone composition is quartzofeldspathic (mean Qm81F18Lt1; S in Figs. 4B and 5C), and it evolves to more mature quartzose sandstones in depocentral areas (mean Qm96F3Lt1; C in Figs. 4B and 5D). This fact suggests an important maturation during transport (∼50 km) in humid climate (Rat, 1982). Quartzofeldspathic petrofacies are indicative of a plutoniclastic origin from coarse crystalline sources from the Hesperian Massif (Arribas et al., 2003). In addition, in the northeast edge of the basin, local supplies from Triassic and Jurassic sedimentary rocks (carbonate and clastics) produce quartzolithic sandstones petrofacies (mean Qm93F1Lt6; N in Fig. 4B). Frame-work replacements by carbonate and kaolinite on K-feldspars have been identified. In addition, few metamorphic lithic grains were crushed producing pseudomatrix. As mentioned earlier, a low-grade metamorphic event (hydrothermalism) took place in depocentral areas of the basin, provoking some mineralogical changes in original sands. Some of these changes are silicification and chloritization of feldspars, metamorphic lithic grains, and intrabasinal argillaceous grains, as well as the growth of chloritoid crystals on these deposits (Barrenechea et al., 1995; Alonso-Azcárate et al., 1999; Mantilla-Figueroa, 1999).
Major-element compositions of the shales and sandstones with derived geochemical parameters and indices are given in Table 1. Absolute concentrations in the different major elements (expressed in oxides) are higher in shales than in sandstones, except for SiO2 (Table 1; Fig. 6).
Sandstones show an intermediate to high content in SiO2, generally between 76.49% and 89.47% in Permian-Triassic (rift 1) sandstones and between 61.13% and 96.27% in Lower Cretaceous (rift 2) sandstones, and both can be considered as mature sandstones (Pettijohn et al., 1973). In some cases the SiO2 content is extraordinarily low (57.85%, sample ALA–10, Table 1) due to pervasive carbonate cementation. In comparison, shales show a typical intermediate content in SiO2 (between 51.98% and 76.63% in rift 1 shales and between 29.74% and 78.35% in rift 2 shales).
Both shales and sandstones have an intermediate Al2O3/SiO2 ratio (0.03–0.39 in rift 1 deposits and 0.02–0.50 in rift 2 deposits). This ratio in sandstones (0.09 in rift 1 sandstones and 0.06 in rift 2 sandstones) shows values lower than shales (0.27 and 0.35 in rift 1 and rift 2, respectively). The high clay minerals content in shales is reflected by a high percentage in Al2O3 (8.88%–21.57% in rift 1 shales and between 8.38% and 26.47% in rift 2 shales; Table 1). Lower values in Fe2O3 + MgO are found in sandstones (1.69–4.08, mean 2.44 in rift 1; 1.50–20.77, mean 6.53 in rift 2) while shale values are higher (4.04–12.66, mean 8.22 in rift 1; 3.07–26.52, mean 16 in rift 2).
Harker diagrams for SiO2 versus Al2O3 show a negative correlation (Fig. 6). Other negative correlations can be observed between SiO2 versus Fe2O3, K2O, MgO, Na2O, and TiO2. Nevertheless, marked positive correlations exist for Al2O3 versus K2O (correlation coefficient r = 0.86 and 0.95 in rift 1 and rift 2 deposits), Al2O3 versus MgO (correlation coefficient r = 0.77 and 0.76 in rift 1 and rift 2 deposits), and Al2O3 versus TiO2 (correlation coefficient r = 0.69 and 0.90 in rift 1 and rift 2 deposits) (Fig. 6).
CaO and MnO versus SiO2 do not present a clear correlation. A CaO-SiO2 scatter could result from carbonate cementation during diagenesis (Gu et al., 2002). Most samples present low CaO percentages (<1 wt%), though in some cases (ALA-10 sample, Table 1), high content is observed (near 16 wt%) due to the presence of carbonate cement. On the other hand, Al2O3 versus CaO presents a negative correlation (correlation coefficient r = −0.30 and −0.18 in rift 1 and rift 2 deposits) in shales because Ca is leached during chemical weathering (Nesbitt et al., 1980).
Large Ion Lithophile Elements (LILEs)
The contents of Rb (30–242, mean 156 ppm in rift 1 deposits; 3–262, mean 106 ppm in rift 2 deposits), Cs (30–119, mean 15 ppm in rift 1 deposits; 0.60–29, mean 10 ppm in rift 2 deposits), Ba (25–13, mean 1144 ppm in rift 1 deposits; 19–1450, mean 368 ppm in rift 2 deposits), and Sr (29–385, mean 130 ppm in rift 1 deposits; 4–246, mean 81 ppm in rift 2 deposits) show a considerable scatter (Table 2) 202, but their average values are comparable with the North American shale composite (NASC) (Gromet et al., 1984) or Average Upper Crust (AUC) (Taylor and McLennan, 1981). On the whole, shales display higher Rb, Cs, Ba, and Sr contents than sandstones.
Positive correlations are observed in K versus Rb (correlation coefficient r = 0.31 and 0.97 in rift 1 deposits and rift 2 deposits, respectively), K versus Sr (correlation coefficient r = 0.31 and 0.38 in rift 1 deposits and rift 2 deposits, respectively), K versus Cs (correlation coefficient r = 0.25 and 0.90 in rift 1 deposits and rift 2 deposits, respectively), and K versus Ba (correlation coefficient r = 0.31 and 0.78 in rift 1 deposits and rift 2 deposits, respectively). This suggests that K-rich clay minerals (such as illite) control the presence of these trace elements (McLennan et al., 1983; Feng and Kerrich, 1990). Sr content is low in rift 1 and rift 2 deposits, which would imply that the source rocks were poor in plagioclase (Feng and Kerrich, 1990).
Rare Earth Elements (REEs)
In Table 3 302, REE compositions are shown for shales and sandstones in each stratigraphic section. La to Lu elements were considered to determine absolute concentrations and several characteristic parameters (ΣREE, ΣLREE, ΣHREE, ΣLREE/ΣHREE, Eu/Eu*, La/Yb, La/Sm, Gd/Yb, where L is light and H is heavy.).
Samples generally show uniform and similar values to the NASC (Gromet et al., 1984) and the AUC (Taylor and McLennan, 1981). The most significant signatures are: (1) high LREE (La to Sm), (2) moderate HREE (Gd to Lu), and (3) a negative Eu anomaly. Values for shales are slightly higher than for sandstones, as seen in the chondrite-normalized REE diagrams (Fig. 7). The general tendency between sandstones and shales and among all the stratigraphic sections is very similar in all diagrams.
The enrichment in LREE in both rift 1 and rift 2 deposits is reflected by a high ratio of (La/Yb)n (7.03–13.22, mean 9.93 and between 4.94 and 17.70, mean 11.34, respectively), (La/Sm)n (3.27–4.21, mean 3.78 and 2.66–4.50, mean 3.79, respectively), and ΣLREE/HREE (6.86–11.76, mean 8.92 and 6.94–15.48, mean 9.70, respectively). A significant negative Eu anomaly (Eu/Eu*) is marked in the diagrams with values between 0.52 and 0.77, mean 0.63, in rift 1 deposits and 0.09–0.93, mean 0.49, in rift 2 deposits.
Weathering of the Source Area
The positive correlations between Al2O3 versus K2O, Al2O3 versus MgO, and Al2O3 versus TiO2 (Fig. 6) indicate that weathering was an important control in the source area in rift 1 and rift 2 deposits (Feng and Kerrich, 1990). Al and Ti are stable or residual elements during chemical weathering, while K and Mg are fixed in the clay minerals and Ca is leached (Nesbitt et al., 1980). Sandstones and shales in rift 1 and rift 2 deposits show variable degrees of negative correlations for SiO2 versus Al2O3 related to the increase of mineralogical maturity (Bhatia, 1983; Gu et al., 2002). These data are supported by the petrographic observations (Fig. 5).
The positive correlations between K versus Ba, K versus Rb, and K versus Cs indicate a clear relationship with alteration of minerals enriched in K (McLennan et al., 1983; Feng and Ker-rich, 1990). Illite is enriched in K and Al, and as a consequence, an increase in illite corresponds to a greater abundance of Al, K, and their positively correlated elements.
The chemical index of alteration (CIA) of Nesbitt and Young (1982) was calculated to estimate the degree of weathering of source rocks (Table 1; Fig. 8). Values for sandstones vary between 63 and 73 in rift 1. Values in rift 2 sandstones are generally higher (71–91). These CIA values are according to petrographic compositional data from both rift sediments (Fig. 5). A better preservation of K-feldspar during the Triassic is observed. This fact may reflect a more intense weathering during Early Cretaceous (rift 2) than in Triassic (rift 1) times, as a consequence of climate conditions (Rat, 1982).
Rift 1 shales present higher CIA values than sandstones (70–78), suggesting that the main weathering products concentrate in shales as clay minerals. Graphic expression of the CIA suggests that weathering produces the alteration of K-feldspar to clay minerals as illite (Fig. 8A). Lower values in rift 1 sandstones can be explained as an intermediate value between those estimated for the unweathered source area (idealized value for granites, see Fig. 8A) and the shales values, outlining the weathering sequence.
In spite of these observations, the presence of hydrothermal chlorite minerals in rift 2 deposits could have increased the content of Al2O3 and biased interpretations about weathering (Grigsby, 2001). In addition, diagenetic alteration of K-feldspars (epimatrix) in rift 1 and rift 2 sandstones (Arribas, 1987; Mata, 1997) would increase the illite content in sandstones and thus increases in Al2O3 and CIA values can be expected.
Anomalous values consisting of high CaO + Na2O content are detectable in some rift 2 sandstones in the northern area of Cameros Basin (TRE-8, TRE-5, and SPM-3; Table 1; Fig. 8B). This can be explained by the nature of source rock (Triassic and Jurassic sedimentary carbonates) and interstitial carbonate cements (Ochoa et al., 2004). Furthermore, sedimentary provenance for samples in this area produces supplies with low contents in siliciclastic minerals.
Several authors (Taylor and McLennan, 1985; McLennan et al., 1995; Gu et al., 2002) have used the Th/U ratio to decipher the weathering history due to the oxidation and loss of uranium during the weathering process. Sediments of rift 1 deposits show a cluster slightly above the upper crust value, with a short weathering trend in shales (arrows in Fig. 9A). This could mean that weathering conditions were constant during Permian-Triassic sedimentation. On the other hand, rift 2 sandstones display anomalous low ratios of Th/U with low Th and U contents, and they plot below the upper crust mean value (Fig. 9B). These anomalies could be related to provenance imprints, and probably to a coarser-grained texture in some samples that produces an important decrease of Th-rich dense minerals, as discussed next. Additionally, weathering trends in rift 2 sandstones and shales are visible and correspond to more intense weathering (higher values of Th/U ratios) than in Permian-Triassic deposits.
Tectonic Setting and Nature of Source Rocks
Using major elements, Bhatia (1983) and Roser and Korsch (1986) determined broad tectonic settings: oceanic and continental island arcs, and active and passive continental margins. According to these authors, both rift 1 and rift 2 deposits plot in or near the passive continental margin field (Fig. 10). Intracra-tonic grabens (aulacogens, e.g., Iberian Basin) are similar to passive-margin settings in terms of the nature of the crust (Bhatia and Crook, 1986).
Trace elements such as La, Th, Sc, Co, and Zr are transferred into clastic sediments during primary weathering, due to their low mobility. Thus, they are useful tools for provenance and tectonic discrimination (Bhatia, 1985; Taylor and McLennan, 1985; McLennan and Taylor, 1991; Bhatia and Crook, 1986; Gu et al., 2002).
We used La-Th-Sc, Th-Co-Zr/10, and Th-Sc-Zr/10 discrimination plots of Bhatia and Crook (1986) to characterize the tectonic setting (Fig. 11). In the La-Th-Sc ternary diagram, rift 1 and rift 2 samples plot in the field defined by Bhatia and Crook (1986) for graywackes from continental margins. However, no discrimination between passive and active continental margins can be observed in this diagram. Shales show a higher content in Sc than sandstones due to the increase in clay minerals where this element is fixed (Mata et al., 2000). In Th-Sc-Zr/10 and Th-Co-Zr/10 ternary plots, rift 1 and 2 sandstones plot mainly in the continental passive-margin field (Bhatia and Crook, 1986), but shales and some rift 2 sandstones plot in the continental island-arc field. This fact is again possibly due to the higher content of Sc in clay minerals. In addition, rift 2 sandstones from the central and northern area of the Cameros Basin show high contents in Zr and Co. The high contents in Zr could be associated with sorting and maturation processes during transport (McLennan et al., 1993; García et al., 2004; Whitmore et al., 2004). High content in Co can be related to provenance, as will be discussed in the following sections.
In rift 1 and 2 deposits, enrichment in LREEs, the characteristic negative Eu anomalies, and the flat HREE patterns suggest derivation from an old upper continental crust composed chiefly of felsic components (Gu et al., 2002).
Floyd and Leveridge (1987) proposed the La/Th versus Hf diagram to discriminate between different source compositions. Most rift 1 and 2 samples plot in the felsic source field (Fig. 12). However, low contents in Hf in some rift 2 samples force them to be plotted in the andesitic field (Fig. 12B).
La/Co average ratios are 10.25 and 6.63 for rift 1 and 2 sandstones, respectively (Table 2) 302. On the other hand, Th/Co average ratios are 3.02 and 1.67 for rift 1 and 2 samples, respectively (Table 2 202; Fig. 13). Sands derived from granitoid sources show higher La/Co and Th/Co values than those sands derived from basaltic sources (Cullers and Berendsen, 1998). Rift 1 and rift 2 sandstones plot in an average range corresponding to sediments derived from upper continental crust.
All these general inferences about tectonic setting and nature of source rocks using minor elements in rift 1 and rift 2 deposits agree with an upper continental crust main provenance. However, some important anomalies are observed specially in some Cretaceous sandstone from the central and northern area of the basin. These anomalies are: (1) high content in Sc, Co, and Zr; (2) low content in Hf, Th, and U, and (3) anomalies in ratios like Th/Y (mean 2.9), La/Tb (mean 35.53), Ta/Y (mean 1.11), and Ni/V (mean 0.72). These data produce important shifts in diagnostic provenance diagrams to intermediate-basic source fields (e.g., continental island arc, Bhatia and Crook, 1986; andesitic source, Floyd and Leveridge, 1987; basalt, Cullers and Berendsen, 1998). However, a clear compositional elemental spectrum that characterizes basic sources is observed, suggesting a mixture of a main felsic source with minor contribution from a mafic source (e.g., alkaline intermediate magmatism). The mafic sources must be considered in relation with a post-Buntsandstein magmatic activity, because these anomalies are not observed in the Triassic deposits (rift 1). Along the northern and southern margins of the Iberian Rift system, and during Norian to Bajocian times, several episodes of basaltic volcanic activity occurred (Salas and Casas, 1993; Martínez-González et al., 1996; Salas et al., 2001; González Menéndez and Suárez, 2005).
The presence of anomalous samples in the northern and central area of the Cameros Basin suggests that basic sources could be associated with Triassic and Jurassic sedimentary rocks located to the north of this basin that acted as sources during Early Cretaceous times.
In summary, petrography and geochemical data suggest that during the first stage of rifting (Permian-Triassic) sources were related to felsic coarse-grained rocks, associated with upper continental crust provenance (Hesperian Massif). In the second stage of rifting, during the most active phase (Berriasian–early Aptian), felsic upper crust provenance was maintained and located in the SW part of the basin (Arribas et al., 2003). In addition, in the NE part of the basin, contemporaneous supplies from a sedimentary cover (Permian-Triassic and Jurassic) were found in association with a volcanic imprint.
During the most active stages of rifting in the intracratonic Iberian Basin (rift 1: Permian-Triassic; rift 2: Late Jurassic to early Albian), quartzofeldspathic plutoniclastic petrofacies were generated in fluvial-lacustrine environments. Composition of Permian-Triassic sandstones varies in time from very quartzose quartzo-lithic/quartzofeldspathic at the base of the succession (Saxonian facies) to K-feldspar–rich quartzofeldspathic petrofacies at the top (Buntsandstein facies). This suggests sedimentation in arid conditions and poor maturation during transport. Composition of Berriasian–early Albian sandstones shows variations from proximal areas (quartzofeldspathic petrofacies) to depocentral zones of the basin, as a result of maturation during transport in a humid climate. In addition, sedimentoclastic petrofacies are found in the northern part of the basin. Both Permian-Triassic and Lower Cretaceous deposits are related with a provenance from the Hesperian Massif and its sedimentary cover.
Weathering inferences from geochemical data agree with petrographic deductions. Thus, CIA values in Permian-Triassic sandstones vary between 63 and 73, while in Lower Cretaceous sandstones, these values vary between 71 and 90, reflecting differences in weathering by climate conditions. However, values of CIA can be modified by (1) diagenetic processes in sandstones (illite epimatrix), which increase the Al2O3 content; (2) sedimentary supplies from the source area (increasing CaO + MgO content); and (3) allochemical hydrothermalism, which produces an increase in Al2O3 by the growth of chlorite minerals.
The use of Th/U ratio could describe a short weathering trend in Permian-Triassic deposits due to the persistent arid conditions during sedimentation. In Lower Cretaceous sandstones, weathering trends are more evident, but with very low ratios and low contents of Th and U.
Geochemical data (major, trace, and REE elements) from both rift 1 and rift 2 deposits fit well in most diagnostic diagrams used for tectonic setting and nature of source rocks.
Ratios between major (Al2O3, SiO2, MgO, K2O, Na2O, and Fe2O3) and trace elements (La, Th, Sc, Co, and Zr) are in agreement with data from passive-margin settings, in terms of the nature of the crust (Bhatia, 1983; Bhatia and Crook, 1986; Roser and Korsch, 1986). REE values show an enrichment in LREE, a flat HREE pattern, and the characteristic negative Eu anomaly in both rift 1 and rift 2 deposits, suggesting a derivation from an old upper continental crust of felsic nature.
Anomalies such as (1) high content in Sc, Co, and Zr; (2) low content in Hf, Th, and U, and (3) anomalies in ratios like Th/Y, La/Tb, Ta/Y, and Ni/V in some Lower Cretaceous sediments suggest an additional basic source related to alkaline volcanism during Norian-Hettangian and Aalenian-Bajocian times. These volcanic sources could be related to the sedimentary cover (Permian-Triassic and Jurassic) located to the north of the Cameros Basin.
Finally, geochemical composition of rift deposits has manifested to be a useful and complementary tool to petrographic deductions, especially in throwing light on provenance from highly weathered sediments under different climate conditions and maturation during transport. However, many processes affecting the original detrital deposits (e.g., diagenetic processes, hydrothermalism) may produce changes in composition that could bias the provenance and weathering deductions.
The authors thank E. Garzzanti and K. Sircombe for their comments and suggestions in a detailed review of the manuscript. We are grateful to M. Muñoz, M.J. Huertas, C. Villaseca, and J. Escuder for helpful suggestions and their remarks about geochemical data, which consistently improved an early version of the manuscript. Funding for this research was provided by the Spanish Government research projects BTE2001-026 and CGL2005-07445-C03-02.
Figures & Tables
Sedimentary Provenance and Petrogenesis: Perspectives from Petrography and Geochemistry
- chemical composition
- clastic rocks
- extension tectonics
- felsic composition
- fluvial environment
- geochemical anomalies
- Iberian Massif
- Iberian Mountains
- Iberian Peninsula
- Iberian Plate
- intracratonic basins
- Lower Cretaceous
- Lower Triassic
- major elements
- Permian-Triassic boundary
- rare earths
- rift zones
- sedimentary rocks
- Southern Europe
- stratigraphic boundary
- trace elements
- Upper Permian
- Variscan Orogeny