Continental geochemical signatures in dacites from Iceland and implications for models of early Archaean crust formation
Introduction
Early Archaean tonalite–trondhjemite–granodiorite (TTG) assemblages provide a unique but genetically little understood geological record of continental crust formation dating back more than 3.8 Ga years (Barker and Arth, 1976, Condie, 1981, Jahn et al., 1981, Martin, 1987, Kröner and Layer, 1992, de Wit, 1998, Martin, 1999, Smithies, 2000). Before plausible geological models for the origin of early felsic continental crust on Earth can be discussed, the processes delivering the key-chemical characteristics of early Archaean TTG assemblages, such as calc-alkaline affinity (low FeO⁎/MgO ratios), enrichment in Na over K, high Ta/Nb ratios, and in particular, positive Pb and negative Nb anomalies in primitive mantle-normalised trace element plots need to be understood (Drummond and Defant, 1990, de Wit, 1998, Martin, 1999, Smithies, 2000, Martin and Moyen, 2002, Smithies et al., 2003).
The chemical similarity of modern adakites (high-Mg andesites) with Archaean TTG rocks suggests that the latter were formed by melting of subducted oceanic crust in a modern-style subduction zone setting (Drummond and Defant, 1990, Martin, 1999, Martin and Moyen, 2002, Kelemen et al., 2004). This is supported by high Ni and MgO concentrations in late Archaean and Proterozoic TTG assemblages that have been ascribed to the interaction of felsic melts with the mantle wedge during ascent (Martin and Moyen, 2002, Smithies et al., 2003). While modern-style plate tectonic processes may have operated since the late Archaean ( < ca. 3.2 Ga years; de Wit, 1998, Smithies et al., 2005, Smithies et al., 2007b, Van Kranendonk, 2007), structural evidence for plate subduction prior to ca. 3.2 Ga years remains equivocal (Kröner and Layer, 1992, Smithies, 2000, Van Kranendonk et al., 2007b, Van Kranendonk, 2007). In fact, low Ni and MgO concentrations in early Archaean TTG assemblages ( > 3.2 Ga) question their formation by such a process (Martin and Moyen, 2002, Smithies et al., 2003).
It has also been suggested that early Archaean TTG assemblages formed by melting of hydrated basaltic plateaus in an ancient plume-dominated environment (Kröner and Layer, 1992, Hamilton, 1998, Martin, 1999, Smithies, 2000, Kemp and Hawkesworth, 2003, Hawkesworth and Kemp, 2006, Bédard, 2006, Smithies et al., 2007a, Van Kranendonk et al., 2007a, Van Kranendonk et al., 2007b, Hansen, 2007). Testing this hypothesis with geological and geochemical evidence in Archaean terrains faces the problem of being circumstantial, as it is not possible to unambiguously determine the geological context in which the earliest TTG crust were formed. In particular, it cannot be precluded that many Archaean TTG associations sampled today were formed by re-melting of older TTG-like continental crust (e.g., Champion and Sheraton, 1997, Bédard, 2006, Van Kranendonk et al., 2007b). Such a process would obscure most if not all evidence related to the initial crust formation event.
Phanerozoic rocks with negative Nb–Ta–Ti and positive Pb anomalies in primitive mantle-normalised trace element plots are assumed to be restricted to volcanic arc settings (e.g., Pearce and Peate, 1995). Therefore, the characteristically high Pb/Ce, Ta/Nb, and La/Nb ratios in Archaean TTG assemblages challenge the validity of the plateau model. Experimental data and trace element modelling suggest that these key-chemical features of Archaean TTG assemblages and the continental crust can form by melting of a basaltic protolith outside a subduction zone environment (e.g., Rapp et al., 2003, Bédard, 2006). Yet, TTG-like rocks are so far unknown in modern oceanic plateaus. White et al. (1999) reported tonalitic rocks with high La/Nb ratios from the Aruba batholith in the south of the Caribbean plateau. However, field relationships suggest that these rocks were formed during collision of the Caribbean plateau with the South American continent (White et al., 1999). In another study Weis et al. (2001) and Ingle et al. (2002) reported continental rocks within the basaltic basement of the Elan Bank (Kerguelen Plateau). Yet, these rocks are not genetically linked to the Kerguelen plateau but rather represent crustal fragments of the Indian continental lithosphere (Ingle et al., 2002).
Felsic samples with tholeiitic and calc-alkaline composition have also been reported from the Króksfjördur volcanic complex, NW Iceland (Pedersen and Hald, 1982, Jónasson et al., 1992). Here, we present new chemical and isotopic data for tonalitic dacites from this locality that demonstrate the striking similarity to those of Archaean TTG assemblages. The Króksfjördur samples hence offer a rare opportunity to decipher the geochemical processes relevant for TTG formation in the early Archaean.
Section snippets
Sample collection and analytical methods
The Tertiary basaltic to felsic Króksfjördur volcanic complex in NW Iceland (Fig. 1) probably developed within a swarm of fissures close to the Icelandic rift about 10 Ma ago (Pedersen and Hald, 1982, Jónasson et al., 1992). Dacitic samples were collected from four localities (Fig. 1). Samples from the Geitafell area are massive, black pitchstones containing phenocrysts of plagioclase and augite in a devitrified glass matrix. At Valshamar and Stryta, two dacitic dome-shaped lava flows were
Comparison of samples with Archaean TTG assemblages
In Fig. 2, Fig. 3, the major- and trace element data of the dacites are compared with those of Archaean TTG associations. Archaean TTG assemblages are calc-alkaline, SiO2-rich (> ca. 65 wt.%; Fig. 2b) rocks. When compared to most post-Archaean granitoid rocks, they are distinctly enriched in Na relative to K (Fig. 2a; Barker and Arth, 1976, Martin, 1987) and can be subdivided into low-Al2O3 (< 15 wt.%) and high-Al2O3 (> 15 wt.%) varieties (Barker and Arth, 1976, Drummond and Defant, 1990). The
Melt contamination by Icelandic mafic crust
It may be argued that some of the plagioclase, clinopyroxene, and amphibole crystals in the dacites may be xenocrysts originating from disintegrated mafic crustal xenoliths. In this case, the mafic crustal xenoliths might have contaminated the dacitic melts. CaO is a major constituent of plagioclase and clinopyroxene, the two major mineral phases in the xenoliths. In a diagram of CaO concentrations versus Nd isotopes (Fig. 4a) the dacites (CaO ca. 4 wt.%) and the mafic crustal xenoliths (CaO
The Króksfjördur dacites – an analogue for early continental crust formation?
Starting with the late Archaean ~ 3.2 Ga years ago, structural and geological evidence, such as the presence of eclogite to blueschist facies rocks, accretionary complexes and fossil suture zones in Archaean terranes are strong evidence for continental crust formation at convergent plate margins (de Wit, 1998, Smithies et al., 2005, Smithies et al., 2007b). Recent Pb–Hf–O isotope data for Archaean zircons suggest that cratonisation and differentiation of a mantle-derived basaltic proto-crust
Summary and conclusions
Dacites from the Cenozoic Króksfjördur volcanic complex are evolved calc-alkaline rocks with high Na concentrations, high Ce/Pb, La/Nb, Ta/Nb ratios, and low Sr concentrations unlike other felsites from Iceland. Their mantle-like Pb isotopic composition precludes involvement of older continental lithosphere in their petrogenesis.
Low FeO⁎/MgO ratios, a limited variation in CaO/Al2O3 ratios, as well as high Pb/Ce, La/Nb, and low Nb/Ta ratios are consistent with a source composed of plagioclase +
Acknowledgements
This work was funded by the German Research Council (DFG) grant He 1857/12 to E. Hegner and by a Max-Planck-Gesellschaft fellowship awarded to M. Willbold. We thank R. Rudnick, T. Elliott, H. Smithies, and S. Müller for discussions and comments on earlier versions of the manuscript. K. Jónasson helped planning the fieldwork and supported us on Iceland. The authors would like to thank H. Smithies and an anonymous reviewer for very thorough and constructive reviews.
References (59)
A catalytic delamination-driven model for coupled genesis of Archean crust and sub-continental lithospheric mantle
Geochim. Cosmochim. Acta
(2006)- et al.
Geochemistry and Nd isotope systematics of Archean granites of the Eastern Goldfields, Yilgarn Craton, Australia: implications for crustal growth processes
Precambrian Res.
(1997) - et al.
Geochemistry of Paleoarchean granites of the East Pilbara terrane, Pilbara craton, Western Australia: implications for early Archean crustal growth
- et al.
Structure of the crust and uppermost mantle of Iceland from combined seismic and gravity study
Earth Planet. Sci. Lett.
(2000) On Archean granites, greenstones, cratons and tectonics: does evidence demand a verdict?
Precambrian Res.
(1998)Archean magmatism and deformation were not products of plate tectonics
Precambrian Res.
(1998)- et al.
Experimental determination of the partitioning behavior of rare earth and high field strength elements between pargasitic amphibole and natural silicate melts
Geochim. Cosmochim. Acta
(2000) - et al.
REE geochemistry and isotopic data of Archean silicic volcanics and granitoids from the Pilbara Block, Western Australia: implications for the early crustal evolution
Geochim. Cosmochim. Acta
(1981) Chemistry and distribution pattern of recent basaltic rocks in Iceland
Lithos
(1972)Silicic volcanism in Iceland: composition and distribution within the active volcanic zones
J. Geodyn.
(2007)