Revisiting the age and paleomagnetism of the Modipe Gabbro of South Africa
Introduction
The transition from the Archean to the Paleoproterozoic (3.0–2.5 Ga) is a key period in Earth's geophysical and geochemical evolution. In particular, this period encompasses an episode of substantial continental crustal growth and preservation, thought to be related to periodic mantle overturn events and extensive plume volcanism (Condie, 2001). This reconstitution of the upper mantle by deeper hotter mantle was followed by, or was synchronous with, the assembly and breakup of some of the planet's earliest supercontinents or supercratons (Nance et al., 1986, Williams et al., 1991, Young et al., 2001, Cheney, 1996, Aspler and Chiarenzelli, 1998, Condie, 2001, Barley et al., 2005), and the development of oxygenated conditions in the atmosphere. Establishing links between these major environmental changes and the suspected mantle-plume tectonic regime that characterized the late Archean is an ongoing research focus that is often hampered by the absence of reliable paleomagnetic and geochronologic constraints.
One of the earliest and most influential paleomagnetic studies on Archean rocks was that of Evans and McElhinny (1966), who examined the Modipe Gabbro in the Kaapvaal craton. This intrusive unit outcrops as a chain of low rising hills, or inselbergs, approximately 15 km long extending from southern Botswana into northern South Africa (Fig. 1). The Modipe Gabbro was emplaced as a sub-horizontal, sill-like body and the low-lying areas between the gabbro inselbergs consist of soils derived from the Gaborone Granite (McElhinny, 1966). Although exposures between the gabbro and the granite are exceedingly rare, Evans and McElhinny (1966) reported a sheared contact between the two lithologies that suggested that the gabbro was intruded by the granite. Importantly, they reported a positive baked contact test between the two lithologies. They also noted exceptionally stable magnetizations in the gabbro with high coercivities (≥180 mT) and high unblocking temperatures (>500 °C) and identified elongate magnetite inclusions exsolved within clinopyroxene as the main carrier of remanence (Evans et al., 1968). Subsequent work on similar occurrences of magnetite exsolved from clinopyroxenes has shown that they are robust recorders of the geomagnetic field (e.g., Renne et al., 2002, Feinberg et al., 2005, Muxworthy et al., 2013). However, these exsolved magnetite crystals are not ubiquitous throughout the Modipe Gabbro, and identical, stable remanences are found in those parts of the intrusion that do not contain them. There is, therefore, at least one other primary magnetic phase present, and we have identified it as part of this study. There are no clear paleohorizontal indicators within the Modipe Gabbro, which precludes evaluating whether a tilt-correction to paleomagnetic directions is needed. Evans and McElhinny (1966) established a paleolatitude of 80° for the Modipe Gabbro without using any form of tilt correction, and the corresponding virtual geomagnetic pole (VGP) calculated from the mean of their sampling sites (33.3° S and 31.1° E) is still used when discussing the tectonic history of the Kaapvaal craton. The exact position of the Kaapvaal craton relative to that of the Pilbara craton is especially important when considering the configurational history of the supercontinent Vaalbara.
Two years after their paleomagnetic analysis of the Modipe Gabbro, McElhinny and Evans (1968) estimated the strength of the Earth's ancient geomagnetic field, or paleointensity, at the time the unit was emplaced. The Earth's magnetic field is generated by convection of the liquid iron outer core and recordings of the strength of the field through time have been highly sought after to address questions about the stand-off distance of the solar wind (Tarduno et al., 2010) and the timing of the nucleation of solid inner core (Smirnov et al., 2003). This study has been updated by the recent work of Muxworthy et al. (2013) to indicate that magnetic field strength at the time of emplacement was approximately 80% of the modern field. However, the usefulness of this paleointensity estimate has been blunted due to ambiguity surrounding the Modipe Gabbro's precise paleolatitude and age.
Until now, the age of the Modipe Gabbro has been roughly constrained by Rb–Sr and K–Ar ages determined by McElhinny (1966). Rb–Sr total rock measurements yielded an age of 2630 ± 470 Ma. K–Ar measurements of pyroxene separates from three sites yielded ages ranging from ∼2600 to ∼3000 Ma with uncertainties of ±100 Ma. The average of these pyroxene K–Ar dates is 2758 ± 126 Ma when recalculated with the calibration of Renne et al. (2011) (Supplemental Data File). K–Ar measurements on plagioclase separates yielded anomalously young ages of 1983–1934 Ma, which were interpreted to possibly reflect thermal resetting by the nearby Bushveld Complex. The average K–Ar plagioclase age of McElhinny (1966), recalculated per Renne et al. (2011), is 1961 ± 52 Ma, which overlaps with integrated 40Ar/39Ar cooling ages for plagioclase from a Bushveld gabbro reported by Cassata et al. (2009). Thus, while portions of the paleomagnetic analysis of the Modipe Gabbro have long been known to be of high quality, it has been difficult to make use of these analyses owing to the poor geochronologic constraints on the unit.
Here we report the results of a new paleomagnetic and geochronologic study of the Modipe Gabbro. We make use of modern paleomagnetic protocols, including least-squares fitting and great circle analyses, to examine and refine the earlier paleomagnetic work of Evans and McElhinny (1966). Rock magnetic techniques are used to better characterize the extent of mineralogic alteration experienced by the magnetic oxides within the gabbro since their formation in the late Archean. The precise emplacement age of Modipe Gabbro is determined using U–Pb TIMS geochronological methods using baddeleyite as a geochronometer. These improved paleomagnetic and geochronologic data allow us to better constrain the precise position of the Kaapvaal craton at the time of the Modipe Gabbro's emplacement and to use the data from the Modipe Gabbro in reconstructions and evaluations of the supercontinent Vaalbara.
Section snippets
Petrography of the Modipe Gabbro
The Modipe Gabbro consists primarily of medium-grained labradoritic plagioclase and clinopyroxene, both of which exhibit progressively increasing levels of alteration toward the southeast. This spatial trend in degree of alteration, combined with the K–Ar plagioclase ages, is consistent with thermal effects of the Bushveld Complex, which outcrops 20 km to the south of the Modipe Complex. In more heavily altered samples, the clinopyroxene is replaced by actinolite and chlorite with minor amounts
U-Pb Geochronology
A description of the U–Pb isotopic methodology and results for sample MO-3 are given in the Supplemental Data File and the Concordia diagram of Fig. 3. Ten single baddeleyite crystals yield a 207Pb/206Pb age of 2784.0 ± 1.0 Ma (2σ), with an MSWD of 1.7. The uncertainty given does not include decay-constant errors; including them produces an uncertainty of ±7.0 Ma. Using the updated 235U decay constant and uncertainty of Mattinson (2010), the 207Pb/206Pb age is 2779.9 ± 1.0 Ma without including
Rock magnetism
Expanded descriptions of the rock magnetic and paleomagnetic methodologies are included in the supplemental information file.
Paleomagnetism
NRM intensities range from 5.33 × 10−2 to 19.2 Am−1 with a median value of 1.04 Am−1 Alternating field demagnetization of ∼6 cm3 specimens displays a range of demagnetization behavior, with most specimens showing a unidirectional decay to the origin, while others exhibit up to three distinct components of magnetization (Fig. 7). The additional components in some samples are characterized by low coercivities and inconsistent directions, which are inferred to reflect weathering, lightning strikes,
The magnetic carrier of the Modipe Gabbro
Petrographic and rock magnetic observations indicate that the Modipe Gabbro is variably altered, with increasing levels of alteration leading to higher concentrations of low coercivity magnetite. Much of this secondary magnetite was produced during the replacement of Fe–Mg-silicates, such as pyroxene and olivine. Despite this alteration, it is still possible to obtain convincing ChRMs related to the emplacement of the gabbro. Specimens with minimum levels of alteration display univectoral decay
Summary and conclusions
Despite its great antiquity, the Modipe Gabbro offers a clear view into the late Archean. Although rock magnetic and petrographic observations show the Modipe Gabbro to be variably altered, perhaps to lower greenschist facies, the Modipe Gabbro's stable and robust remanence allow the derivation of convincing ChRMs related to the original emplacement and cooling of the gabbro. The Modipe Gabbro's emplacement age of 2784.0 ± 1.0 Ma indicates that it was part of a larger, rapidly emplaced igneous
Acknowledgements
The authors are grateful to M.E. Evans for discussions about the Modipe Gabbro at the beginning of this project and to Stephen Dell of the Madikwe Game Reserve for allowing unrestricted access to the Modipe Complex. Rudy Wenk and Roland Merkle are thanked for field and logistical support, respectively. Lisa Smeenk helped measure paleomagnetic samples at the BGC. Travis Drake measured rock magnetic properties as part of the NSF-Research Experience for Undergraduate program at the University of
References (38)
- et al.
Two Neoarchean supercontinents? Evidence from the Paleoproterozoic
Sedimentological Geology
(1998) - et al.
Late Archean to Early Paleoproterozoic global tectonics, environmental change and the rise of atmospheric oxygen
Earth and Planetary Science Letters
(2005) Cyclic continental mafic tuff and flood basalt volcanism in the late Archaean Nullagine and Mount Jope supersequences in the eastern Pilbara, Western Australia
Precambrian Research
(2001)- et al.
Argon diffusion in pyroxenes: implications for thermochronometry and mantle degassing
Earth and Planetary Science Letters
(2011) - et al.
Argon diffusion in plagioclase and implications for thermochronometry: a case study from the Bushveld Complex, South Africa
Geochimica et Cosmochimica Acta
(2009) Sequence stratigraphy and plate tectonic significance of the Transvaal succession in southern Africa and its equivalent in western Australia
Precambrian Research
(1996)- et al.
Validating the existence of Vaalbara in the Neoarchean
Precambrian Research
(2009) - et al.
Single domain magnetite and high coercivities in a gabbroic intrusion
Earth and Planetary Science Letters
(1968) The paleomagnetism of the main zone in the Western Bushveld Complex
Earth Planetary Science Letters
(1986)The paleomagnetism of the main zone of the Eastern Bushveld Complex
Tectonophysics
(1986)