1
References Dressler, B. O. & Reimold, W.U. 2004. Order or chaos? Origin and mode of emplacement of breccias in floors of large impact structures. Earth-Science Reviews, 67, 1–54. Garde, A.A. & Klausen. M.B. 2016. A centennial reappraisal of the Vredefort pseudotachylytes: shaken, not stirred by meteorite impact. Journal of the Geological Society, London. doi: 10.1144/kgs2015-147 (printed version in November 2016 issue). Garde A. A., McDonald, I., Dyck, B. & Keulen, N. 2012. Searching for giant, ancient impact structures on Earth: the Mesoarchaean Maniitsoq structure, West Greenland. Earth and Planetary Science Letters, 137–138, 197–210. Garde, A. A., Dyck, B., Esbensen, K. H., Johansson, L. & Möller, C. 2014. The Finnefjeld domain, Maniitsoq structure, West Greenland: Differential rheological features and mechanical homogenisation in response to impacting? Precambrian Research, 255, 791–808. Garde, A. A., Boriani, A. & Sørensen, E.V. 2015. Crustal modelling of the Ivrea-Verbano zone in northern Italy re-examined: coseismic cataclasis versus extensional shear zones and sideways rotation. Tectonophysics, 662, 291–311. Gibson, R.L. & Reimold, W.U. 2008. Geology of the Vredefort impact structure: a guide to sites of interest. Council for Geoscience, Pretoria, Memoir, 97, 181 pp. Kreslavsky, M. A. & Head, J. W. 2012. New observational evidence of global seismic effects of basin-forming impacts on the Moon from Lunar Reconnaissance Orbiter Laser Altimeter data. Journal of Geophysical Research, 117, E00H24. Lieger, D., Riller, U. & Gibson, R. L. 2011. Petrographic and geochemical evidence for an allochthonous, possibly impact melt, origin of pseudotachylite from the Vredefort Dome, South Africa. Geochimica et Cosmochimica Acta, 75, 4490–4514. Melosh, H. J. 1979. Acoustic fluidization: a new geologic process? Journal of Geophysical Research, 84, 7513–7520. Melosh, H. J. 2005. The mechanics of pseudotachylite formation in impact events. In: Koeberl, C., Henkel, H. (eds) Impact Studies Series, Springer, Heidelberg, 1, 55–80. Mohr-Westheide, T., Reimold, W. U., Riller, U. & Gibson, R.L. 2009. Pseudotachylitic breccia and microfracture networks in Archean gneiss of the central uplift of the Vredefort impact structure, South Africa. South African Journal of Geology, 112, 1–22. Richardson, J. E., Melosh, H. J., Greenberg, R. J. & O'Brien, D. P. 2005. The global effects of impact-induced seismic activity on fractured asteroid surface morphology. Icarus, 179, 325–349. Shand, S. J. 1916. The pseudotachylyte of Parijs (Orange Free State) and its relation to ‘trap-shotten gneiss’ and ‘flinty crush rock’. Quarterly Journal of the Geological Society, London, 72, 198–217. Sibson, R.H. 1975. Generation of pseudotachylyte by ancient seismic faulting. Geophysical Journal International, 43, 775–794. Spray, J. 1987. Artificial generation of pseudotachylyte using friction welding apparatus: simulation of melting on a fault plane. Journal of structural geology, 9, 49–60. ★ ★ Vredefort Parys 27°00’S Salvamento Leeukop Vaal river island Vaal river island Leeukop Esperanza 10 km Granophyre dyke ★ ★ ★ Salvamento ★ Otavi Vaal V R E D E F O R T C O L L A R VREDEFOR T COLLA R VREDEFORT DOME VREDEFORT DOME 27°00’S 27°30’E Granophyre dyke ★ Johannesburg South Africa Vredefort Pseudotachylyte zone (as mapped by Dressler & Reimold 2004) Observations and samples, this study ★ Pl Qtz Qtz 1 mm 1 mm Pl Pl Qtz 1 mm Qtz Qtz Qtz Kfs Pl Bt Pl Pl Qtz Pl + Kfs Pseudotachylyte matrix consisting of intensely comminuted quartz and plagioclase and tiny laths of K-feldspar. The K-feldspar laths have a preferred planar orientation and wrap around the quartz fragments. This suggests overall compression with orientation lower left – upper right relative to the image frame. Note absence of simple shear textures. The textures are compressional, reflecting pure shear. BSE-SEM image. Interpretation of the microstructures The pseudotachylyte is not a melt in the petrological sense, where melt is produced by chemical reactions between several mineral (± fluid) phases. It is essentially The pseudotachylyte matrix is composed of numerous small, equant to elongate fragments of quartz and plagioclase, besides tiny laths and patches of microcline and minor biotite (both of which may have crystallised from mineral melts). Backscattered electron SEM image and photomicrograph (right part of the same area). a mixture of fragments of intensely comminuted quartz and plagioclase and minute laths and irregular patches of K-feldspar. The latter may have crystallised from K-feldspar melt created by frictional heating to above 1300°C. The observation that pseudotachylyte melting is preceded by cataclasis is not new, as shown in experiments by Spray (1987). Both these experiments and an influential paper by Sibson (1975) assumed that pseudotachylyte formation was caused by frictional heating between sliding surfaces in faults and were controlled by simple shear. However, it was shown by Melosh (2005) that the frictional heating between sliding surfaces will cease as soon as a melt is produced, hence the copious amounts of pseudotachylyte at Vredefort could not have been produced by simple shear along faults as commonly assumed. Furthermore, no major faults have been found in the Vredefort dome where the pseudotachylytes occur. Our interpretation of the Vredefort pseudotachylytes as the results of impact-induced seismic shaking is simple and adequately solves these problems. The comminution and concomitant frictional heating would continue for as long as the seismic shaking persisted – probably for a few minutes after the impact, and at a frequency of around 1 Hz (J.H. Melosh, pers. comm., 2013). This allows for hundreds of violent oscillations during which individual components of the rock would be banging against each other. The main stress component was pure shear, rather than simple shear as previously assumed with presumed fault-bound pseudotachylyte formation. The rapid alternations between strong compressive and dilational stress during seismic shaking would cause very effective breakdown of the host rock and its constituent minerals. Photomicrograph of the same sample collected adjacent to the pseudotachylyte. All feldspar grains are fractured, and original quartz grains are now fine-grained mosaics. Note the absence of penetrative shear strain or overall lateral displacement. The cataclasis is heterogeneous, with a better preserved area in the upper left and a strongly comminuted area in the right of the image. Photomicrograph with crossed polarisers. Enlargement of cataclastic grain of plagioclase from the host rock with combined syn- and antithetic microfaults. Small blocky feldspar fragments have been displaced in opposite directions (arrow). Significance of wall rock observations: Although the major pseudotachylyte zones might appear as distinct veins, there is a gradual transition from host rock that is still coherent but intensely affected by cataclasis and into pseudo- tachylyte. The latter is mainly a finely comminuted rock rather than a melt. In his original definition of pseudotachylyte, Shand (1916) did not imply melting but only aphanitic, almost glassy appearance. Hence the prefix ‘pseudo’. Superficially almost intact orthogneiss adjacent to pseudotachylyte at the Salvamento quarry (sample 524307). On closer inspection, a dendritic microfracture system with beginning cataclasis is seen traversing the centre of the slab (arrows). 0.5 mm 2 mm 2 cm Dendritic fractures and cataclasis Dendritic fractures and cataclasis Pseudotachylyte formation by fric- tional heating along slip surfaces? This is the classical interpretation which goes all the way back to Shand (1916) – who was nevertheless aware of one of its main limitations, namely the absence of the necessary faults. See also the adjacent box ‘Discussion’. ‘Shock’ or ‘flash’ melting? The idea of flash melting by the shock wave itself to form pseudotachylyte has been proposed by several authors including Dressler & Reimold (2004), Gibson & Reimold (2008) and Mohr-Westheide et al. (2009). There are several reasons why this concept does not work. Shock melting takes place at the contact between the impactor and target. Beyond this point the shock wave gives rise to shock melting of individual minerals, as determined by their direct melting temperatures. Furthermore, each of the Vredefort pseudotachylytes is the product of a long series of repetitive events comprising gradual cataclasis and comminution, as shown in this poster. Finally, pseudotachylytes do not occur in the centre of the Vredefort dome, where evidence of the highest shock is recorded. Allochthonous melts? Lieger et al. (2011) made a case for an allochthonous origin of the Vredefort pseudotachylytes from crater floor melts. However, a number of geochemical and isotopic studies, also going all the way back to Shand (1916), have repeatedly shown that the compositions of the pseudotachylytes and their host rocks are the same. Furthermore, excellent candidates for genuine allochthonous melts intruded from above exist in the form of the so-called granophyre dykes, which are well exposed in the southern part of the Vredefort dome. These dykes are genuine intrusive igneous rocks and differ from pseudotachylytes in all respects. Relation to the central uplift? Mohr-Westheide et al. (2009) made a compre- hensive study of microfractures in a polished slab of pseudotachylyte measuring 1.5 by 3 m from an unspecified Vredefort quarry, and attempted to relate the pseudotachylyte formation to the uplift of the Vredefort dome. Unfortunately the size limitation of this study seems to have prevented recognition of the dendritic fracture system and a distinction between tectonic and impact-related joints and fractures. These authors worked from a theory of injection and dilation of the pseudotachylytes, and argued that clasts fit together like pieces in a jigsaw puzzle and were separated by dilation. However, this is not an accurate observation. Only the clasts that have been shaken but not stirred fit nicely together, but the pseudotachylyte mantles of these clasts are thin and without signs of dilation. Clasts with thicker mantles no longer fit together, because they have been both shaken and stirred, diminished in size and rounded in the course of the cataclastic process. Arrested cataclasis and pseudotachylyte formation along shock-induced dendritic fractures in the Salvamento quarry. Note the complete absence of faulting and displacements. Most fractures with cataclasis are discordant to lithological boundaries, others follow them (black arrow). The host rocks are orthogneiss and pegmatite. Arrested cataclasis and pseudotachylyte formation along shock-induced dendritic fractures in the Esperanza quarry. Asterisks mark nodes with progressive formation of pseudotachylyte. Older tectonic joints and a young subhorizontal relaxation joint are also present. Note absence of faulting and displacements beyond a few centimetres. Detail of fracturing and pseudotachylyte formation. The white asterisks mark sites with diminishing and rounding of individual small blocks. Note: The thickness of the pseudotachylyte corresponds to the rounding and size reduction of the remaining blocks. The pseudotachylyte zones were developed at the site itself from the local host rocks. Full-blown pseudotachylyte zone in the Salvamento quarry. Note sharp-edged blocks with thin linings of pseudotachylyte (’Shaken’) and rounded blocks surrounded by much thicker pseudotachylyte (’Shaken and stirred’). The copious zones of pseudotachylyte were developed at the site itself from the local host rocks. Previous studies have shown that the compositions of the local host rock, the clasts and the pseudotachylyte are almost or completely identical. Shock-induced dendritic fracturing with beginning cataclasis in the Esperanza quarry (see the map for location). The host rock in this quarry is granite. Asterisks mark nodes with progressive formation of pseudotachylyte. Note near-absence of displacement along the fractures (as recorded by the thin oblique pegmatite vein left). ★ ★ ★ 1 m ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ 20 cm ★ ★ 50 cm Shaken Shaken and stirred Shaken and stirred Shaken Impact-induced dendritic fracture system Impact-induced dendritic fracture system Impact-induced dendritic fracture system Relaxation joint Tectonic joints Tectonic joint Tectonic joint 20 cm Shaken and stirred Shaken Impact-induced dendritic fracture system Tectonic joint Tectonic joint Pegmatite ★ 1 m ★ Progressive formation of pseudotachylyte in situ Host rocks were also affected by the cataclasis Microstructures: mainly cataclasis, not full melting Some previous models and their shortcomings Discussion • Major pseudotachylyte zones constitute a spectacular component of the renowned, c. 2.023 Ga Vredefort impact structure in South Africa, but it has always been difficult to explain how they were formed. • In his famous original account of the pseudotachylytes, S. J. Shand (1916) interpreted them as due to cataclasis and frictional heating. Shand [v] also pointed out two enigmas which have remained unsolved: • 1) Shand realised that the volumes of pseudotachylyte he observed at Vredefort greatly exceeded those of other pseudotachylytes located within faults elsewhere on Earth, • 2) At Vredefort no associated major faults were identified. • Today we know that the Vredefort pseudotachylytes are impact-related, but how were they actually formed? • The investigation of the Vredefort pseudotachylytes was prompted by the discovery of multiply repeated cataclasis by intense, impact-induced seismic shaking in the Maniitsoq structure, West Greenland (observations in 2010, 2011 and 2016, Garde et al. 2012, 2014). • Further inspiration was provided by observations of endogenic, earthquake-induced seismic shaking in a 45 km long linear belt in the footwall of the Insubric Line, Southern Alps in Italy (Garde et al. 2015). Background and objectives Conclusions – see also Garde & Klausen (2016) Seismic shaking: an important but overlooked cratering process • Observations in the host rocks are the key to understand the pseudotachylytes. Their development was initiated by a shock-induced, dendritic fracture system that penetrated the granitic host rocks and temporarily loosened individual blocks with sizes from centimetres to metres. • After the initial fracturing, intense impact-induced seismic shaking of the now loosened blocks destroyed their margins by cataclasis , whereby the blocks gradually became smaller and smaller and more and more rounded. The cataclasis led to frictional heating and eventually incomplete melting (see box below). • The seismic shaking punched the rock into a very fine powder and heated it by friction, whereby it became the material we call pseudotachylyte. It was produced in situ and not injected from anywhere. • Shocks induce seismic waves, as can be observed in any seismogram. Impact-induced seismic shaking has been described from the Moon (Kreslavsky & Head 2012) and from asteroids (Richardson et al. 2005). Surprisingly, it has been overlooked in the context of terrestrial impacting except in the theoretical concept of acoustic fluidisation of the crater floor (Melosh 1979). Pseudotachylyte, Salvamento quarry, and coauthor Martin B. Klausen Photograph of outcrop with arrested pseudotachylyte formation Distribution of pseudotachylyte, traced from photograph of outcrop Interpretation of further pseudotachylyte formation with continued seismic shaking Full-blown pseudotachylyte zone, Leeukop quarry Shock-induced dendritic fractures, traced from photograph of outcrop (prior to impact-induced seismic shaking) 1 2 3 4 The Vredefort pseudotachylytes – a centennial reappraisal of Shand (1916): shaken, not stirred by meteorite impacting Adam A. Garde 1 and Martin B. Klausen 2 1 Geological Survey of Denmark and Greenland, Copenhagen, Denmark. E-mail: [email protected] 2 Stellenbosch University, Private Bag X1, Matieland 7602, South Africa
