Sediments of two Gondwana glaciations in Ethiopia: Provenance of the Enticho Sandstone and the Edaga Arbi Glacials

• Main Author: Lewin, Anna
• Format: Others
• Language: en
• Published: 2020
• Online Access: https://tuprints.ulb.tu-darmstadt.de/13300/13/Dissertation_AnnaLewin.pdf; https://tuprints.ulb.tu-darmstadt.de/13300/7/Supplementary%20material_dissertation_AnnaLewin.pdf; Lewin, Anna <http://tuprints.ulb.tu-darmstadt.de/view/person/Lewin=3AAnna=3A=3A.html> (2020): Sediments of two Gondwana glaciations in Ethiopia: Provenance of the Enticho Sandstone and the Edaga Arbi Glacials. (Publisher's Version)Darmstadt, Technische Universität, DOI: 10.25534/tuprints-00013300 <https://doi.org/10.25534/tuprints-00013300>, [Ph.D. Thesis]

The Gondwana supercontinent was completely assembled in the Late Neoproterozoic by closure of the Mozambique Ocean and formation of the extensive East African Orogen at the suture of East and West Gondwana. A peneplain formed on the northern margin of the supercontinent, on which a vast blanket of Palaeozoic sediment was deposited. Major amounts of sediment are assumed to have been eroded and transported from the East African Orogen to the continental margin via large sediment fans (Gondwana super-fan system). The Palaeozoic sedimentary succession of northern Ethiopia evidences two Gondwana glaciations, which are recorded in the Upper Ordovician–Lower Silurian Enticho Sandstone and the Upper Carboniferous–Lower Permian Edaga Arbi Glacials. These formations have been studied sedimentologically and palynologically, but their provenance remained unclear. This thesis presents a multi-method provenance study on samples of these two formations. Thin section petrography provides the basis; then the bulk sandstone samples were analysed for their major and trace element geochemistry. Heavy minerals were separated from the samples and their assemblage was determined. Mineral chemical analyses were conducted on rutile and garnet grains from both formations. Finally, detrital zircon ages were determined. The study followed the objective to fill a data gap in correlation of provenance patterns across Palaeozoic sedimentary rocks of northern Gondwana, providing further insights into the Palaeozoic sediment dispersal system and the influence of the two glaciations on sediment provenance. The Enticho Sandstone is composed of tillite at the base followed by glaciogenic sandstone, probably representing meltwater deposits. At the top of the formation, better sorting and distinct cross stratification show shallow marine reworking and evidence the post-glacial transgression. The Edaga Arbi Glacials comprise tillite and finely laminated sand- and siltstone with dropstones, interrupted by sandy layers. The two formations differ strongly in their mineralogical maturity. The Enticho Sandstone is highly mature and unusually quartzose for glaciogenic sandstone. The marine sub-unit shows even higher quartz contents. The geochemical composition underlines this high maturity and yields a high chemical index of alteration (85), pointing to intense chemical weathering and reworking of the material. It is likely that the alteration has taken place before the glaciation. The Edaga Arbi Glacials feature lower maturity with higher amounts of feldspar and rock fragments and a chemical index of alteration of 62. Trace and rare earth elements indicate a higher influence of juvenile source material than for the Enticho Sandstone. Juvenile crustal rocks are abundant in the underlying Nubian Shield. Comparison of the geochemical data with age-equivalent formations in Saudi Arabia shows similar patterns for the Ordovician–Silurian, but major differences in the Carboniferous–Permian, supporting previous assumptions of a large, uniform sedimentary system during the Late Ordovician glaciation and more localised sediment transport during the Carboniferous–Permian. Detrital zircon chronology resulted in main age populations of Pan-African (700–550 Ma), Tonian (900–700 Ma), Stenian–Tonian (1200–900 Ma) and minor Palaeoproterozoic and Archaean zircons for both formations. The relation of the Tonian and Stenian–Tonian populations, however, differs strongly between the two formations. The Enticho Sandstone is characterised by a prominent Stenian–Tonian population, which can be used to trace the Gondwana super-fan system. Correlation with Upper Ordovician (glaciogenic) and Cambrian–Ordovician sandstones in northern Africa and the Middle East yields high similarity with those in Israel, Jordan and Libya, which are assumed to represent a super-fan. It further shows that no change in zircon age patterns occurs with the onset of the glaciation. It is thus likely that the Enticho Sandstone contains recycled super-fan material. The Edaga Arbi Glacials have a characteristic Tonian population. Such ages are omnipresent in the southern Nubian Shield and represent its earliest formation stage, supporting the assumption of a rather proximal provenance. No regional or stratigraphic trends could be observed within one of the studied formations. The heavy mineral assemblages of both formations are highly different as well. The Enticho Sandstone is characterised by a large proportion of the ultra-stable heavy minerals zircon, tourmaline and rutile (ZTR). In the lower, glaciogenic sub-unit, significant amounts of garnet are also present. In the Edaga Arbi Glacials, on the other hand, apatite and garnet make up most of the heavy mineral assemblage. No stratigraphic trends were identified within the Edaga Arbi Glacials. Neither could regional trends be observed in one of the studied formations. These patterns underline the differences in mineralogical maturity revealed by petrography and geochemistry. Very little chemical alteration must have affected the Edaga Arbi Glacials, whereas the material forming the Enticho Sandstone is strongly altered. In the well sorted and permeable marine subunit of the Enticho Sandstone it is likely that diagenetic modification by corrosive pore fluids took place and reinforced the high mineralogical maturity. Rutile and garnet chemical analyses point to a combination of magmatic and metamorphic source rocks with metamorphic temperatures of mainly amphibolite-, but also granulite-facies grade for both formations. For the Enticho Sandstone, the heavy mineral analysis confirms the assumption that it contains recycled super-fan material, which was strongly weathered before on the North Gondwana peneplain. The garnet is thought to have been delivered by varying erosion of the basement of the Saharan Metacraton (and maybe also the Nubian Shield) by the glaciers. The proximal provenance of the Edaga Arbi Glacials is confirmed again by the high amounts of unstable heavy minerals. Since the directly underlying basement does not contain high-grade metamorphic rocks, a provenance from the southern hinterland is likely, where the Nubian Shield merges the Mozambique Belt and higher metamorphic grades were reached. Combining all methods and their outcome with information from the literature, the following provenance models can be inferred for the two studied formations. The Enticho Sandstone was formed during the Late Ordovician (Hirnantian) glaciation, when a large ice sheet covered much of northern Africa with a spreading centre in North-West Africa. The ice reached as far south-east as the study area. Ice and meltwater transported sediments to the study area; sediments, which were spread before via the Gondwana super-fan system and strongly weathered on the North Gondwana peneplain during the Cambrian and pre-glacial Ordovician. The original provenance of this super-fan material remains unclear. Material of the Saharan Metacraton basement was eroded by glaciers and admixed to variable amounts. During the post-glacial transgression in the Early Silurian, the upper part of the sedimentary succession was reworked by seawater without adding new detritus. In contrast, the Edaga Arbi Glacials are sourced from the southern hinterland of the Nubian Shield at the transition to the Mozambique Belt. In the Late Palaeozoic, a complex regional topography led to mountain glaciers that eroded the uplifted basement and transported material to nearby depressions, in which proglacial lakes formed. A period of non-deposition between the two formations may have been caused by a consecutive combination of isostatic-rebound after the Late Ordovician glaciation, eustatic sea-level fall in the late Silurian–early Devonian and up-doming prior to Neo-Tethys rifting. No recycling of the Enticho Sandstone by the Edaga Arbi Glacials took place on a grand scale. This was either because the deposition of the former was limited to northern Ethiopia and the source area for the latter was to the south or because the Enticho Sandstone was eroded in the source area of the Edaga Arbi Glacials before the Carboniferous–Permian glaciation.