Tracing southern Gondwanan sedimentary paths: A case study of northern Namibian late Palaeozoic sedimentary rocks

The northern Namibian Karoo‐aged successions are part of a Gondwana‐wide sedimentary system emerging at the Carboniferous–Permian boundary and existing for more than 50 Ma. The Karoo Supergroup sedimentary successions are of importance in understanding the evolution of the Karoo rift system. This study presents new whole‐rock geochemical data combined with detrital zircon morphology as well as U−Pb ages and Lu−Hf composition of late Palaeozoic siliciclastic rocks of the Namibian Huab Basin and Kunene area (south‐west Africa). Inferred by youngest detrital zircon U−Pb ages the Verbrande Berg Formation (lower Ecca Group) yields a Sakmarian to Asselian maximum depositional age, whereas the overlying Tsarabis Formation yields an Artinskian maximum depositional age. These ages coincide with the end of the Dwyka ice age and an overall warming and a contemporaneous evolution of the Karoo I rift system across southern Gondwana. The zircon age distribution of the investigated samples yields clusters ranging between ca 500 to 650 Ma (Cambrian–late Neoproterozoic), ca 950 to 1200 Ma (early Neoproterozoic–Mesoproterozoic) and ca 1800 to 1900 Ma (Palaeoproterozoic). Their rounded shapes characterize the zircon grains of the Kunene area and the lower Huab Basin section, whereas upper Huab Basin strata yield mostly unrounded grains. The rounded nature of zircon grains with a diverse U−Pb age spectrum putatively points towards sediment homogenization and multiple recycling stages during the deposition of the sediments and large catchment areas of the depositing rivers. As suggested by zircon grains with a low roundness value and a single Palaeoproterozoic age cluster, the upper Huab Basin successions were probably deposited under drier climatic conditions, small catchment areas and limited sedimentary homogenization. Therefore, the southern Gondwana sedimentary transport and homogenization system may change over time and is dependent on the climate prevailing during deposition. This study shows that the laws of detrital zircon are very complex and are yet to be explored.


INTRODUCTION
Detrital zircon is widely used as a provenance indicator for clastic sediments, assuming that the investigated material inherits information of its original (proto)source (Iizuka et al., 2010;Blanco et al., 2011;Iizuka et al., 2013;Roberts & Spencer, 2015), which is commonly considered as 'source-to-sink'. This approach seems very questionable given the robustness and durability of zircon crystals (Fedo et al., 2003;Zoleikhaei et al., 2016). Consequently, sedimentary rocks tend to contain zircon grains that are not exposed within their former drainage area acting as intermediate sedimentary repositories (Pereira & Gama, 2021). Widely unrecognized, a contrasting model describing a Gondwana-wide sediment homogenization system originally proposed by Andersen et al. (2016) may be attributable for major parts of the Carboniferous to Jurassic Karoo Supergroup basin sedimentary record. The laws of such a homogenizing sedimentary system are not known yet, as different sedimentary environments may also cause different homogenization patterns (Andersen et al., 2016). Identifying and interpreting such patterns are of great interest as implications drawn from detrital zircon data are widely used for reconstructing palaeosediment fluxes (Meinhold et al., 2020;Cornell et al., 2021;Gibson et al., 2021).
The late Palaeozoic Karoo Supergroup sediments are widespread across the Gondwana supercontinent ( Fig. 1). Both, the Main Karoo and the Paran a Basin reach a maximum thickness of ca 10 000 m (Visser & Praekelt, 1996) and ca 4000 m (Zal an et al., 1990), respectively. Karoo-aged basins of Namibia feature much thinner successions of <1200 m (Karasburg Basin; Berti, 2015). The least pronounced of these successions is located in the northern part of the country, namely the Huab Basin with a thickness of ca 250 m (Horsthemke, 1992), and crops out along the southern banks of the Kunene River with a thickness of ca 55 m (Frakes & Crowell, 1970). Despite their small area these successions constitute a world-class archive of the Late Palaeozoic Ice Age and its demise (Du Toit, 1921) featuring the glacialinduced Dwyka and deglaciated Ecca groups (Stollhofen et al., 2000b;Miller, 2008) including the first documented pre-Cenozoic fjord network morphologies (Dietrich et al., 2021). Their sedimentation was triggered by several phases of extension and thermal uplift leading towards contrasting stratigraphic records developing within the Gondwana interior (Frizon de Lamotte et al., 2015; and references therein). The resulting graben and half-graben structures of the so-called southern African Karoo I rift system ( Fig. 1; Delvauxi, 2004;Tankard et al., 2009) of Namibia were in large parts filled with Gondwanan peneplain material (Zieger et al., 2019(Zieger et al., , 2020a. Similar sediment patterns have also been observed within the detrital zircon record of Karoo-aged sedimentary rocks of the Main Karoo, Paran a and the Mid-Zambezi basins (Andersen et al., 2016;Canile et al., 2016;Viglietti et al., 2018;Barrett et al., 2020) with an increasing amount of Permian-aged zircon material going upsection (Canile et al., 2016;Zieger et al., 2020a).
The study area offers excellent prerequisites examining the late Palaeozoic Namibian successions with respect to the hypothesis of a sediment homogenization system across southern Gondwana and thereby exploring its influencing factors influencing such a system. This was achieved by studying nine samples from two of the smallest occurrences of Namibian Karooaged successions: The Huab Basin successions and the banks of the Kunene River (referred as 'Kunene section'; Fig. 1). The samples of the fluvial to limnic Permian Verbrande Berg, Tsarabis, Gudaus and Gai-as formations (Horsthemke, 1992) as well as the Kunene section were investigated with respect to their detrital zircon age distribution pattern as well as grain morphology patterns. Revealing a detailed provenance analysis was achieved using laser ablationinductively coupled plasma mass spectrometry (LA-ICP-MS) UÀPb and LuÀHf dating techniques as well as whole-rock geochemical analysis. This approach provides an integrated assessment of different features provided by detrital zircon and may become standard for interpreting detrital zircon data.

Basement evolution
Namibia's geological evolution has a history from the Neoarchean to the Mesozoic involving the assembly and break-up of several supercontinents, Rodinia in the late Mesoproterozoic and Gondwana in the Neoproterozoic to Phanerozoic (Evans, 2009;Torsvik & Cocks, 2011). Today Namibia consists of the Neoarchean to Palaeoproterozoic Congo Craton in the north with the Kamanjab Inlier at its margin and the surrounding Mesoproterozoic to Neoproterozoic belts of the Kaapvaal and Congo cratons, representing the remaining parts of the country (Seth et al., 1998;Jacobs et al., 2008;Van Schijndel et al., 2011;Kleinhanns et al., 2015;Fig. 2). The NamaquaÀNatal Belt, which is related to the formation of Rodinia (Li et al., 2008;and references therein), is located in southern Namibia and borders the Kaapvaal Craton in the south-west (Fig. 2). The Gondwana assembly in the Neoproterozoic (ca 870 to ca 550 Ma; Kr€ oner & Stern, 2005) led to the formation of the pan-African mobile belts around the Mesoproterozoic to Archean cores. In Namibia the Gariep Belt is located in the south-west, the Damara Belt is located in the central part, whereas the Kaoko Belt is located in the north-west (Fig. 2).

South-west Gondwanan rifting history and related sedimentation
Since Cambrian times, subduction on the western margin of Gondwana was active (Ramos et al., 2010), and ended with the final accretion of Patagonia in the Carboniferous to Permian, forming the Gondwanide Orogen during Pangea assembly ( Fig. 2; Ramos & Aleman, 2000;Ramos, 2008). The accretion resulted in an early Permian volcanic arc in northern Patagonia and a Permian to Triassic fold-and-thrust belt (Fig. 2), continuing into Africa as the contemporary Cape Fold Belt (Du Toit, 1921). The breakup of Pangea began by the latest Carboniferous by the development of early rift-systems throughout Gondwana (Frizon de Lamotte et al., 2015; and references therein). The southern African Karoo I rift system (Delvauxi, 2004;Tankard et al., 2009) developed obliquely to and was influenced by the Cape Fold Belt accretion ( Fig. 1; Delvauxi, 2004). The system lasted ca 30 Ma from the late Carboniferous until Permian times . The Karoo I rifts may be invoked by the activation of pre-existing pan-African structures and were not accompanied by magmatic activity (Frizon de Lamotte et al., 2015). Lower units of the Huab Basin successions feature some ash fall deposits, but a lack of datable ash layers in the upper part becomes apparent (Stollhofen et al., 2000a;Zieger et al., 2019). Namibian Karoo-aged Aranos and Karasburg basins as well as the Huab Basin were created due to the latter rifting (Catuneanu et al., 2005). These west-east trending rift basins were eventually connected to the ocean and filled with glacial, glacio-marine as well as marine sediments (Catuneanu et al., 2005) aged strata reflects material, which was recycled multiple times. The sediment dispersal especially within the Karoo sedimentary system may be part of a much older recycling system obscuring the information inherited within the detrital zircon age provenance analysis (Andersen et al., 2018a). Nevertheless, the provenance of sedimentary rocks is still an invaluable tool in order to determine the drainage behaviour of ancient sedimentary systems (e.g. Canile et al., 2016;Zieger et al., 2020a).

The northern Namibian Karoo-aged Huab Basin and Kunene section
The Huab Basin is comprised of ca 300 m thick lacustrine to marine successions of Carboniferous to Permian age ( Fig. 4) located in the northwest part of Namibia. The three main outcrop areas are the Goboboseb Mountains area in the north, the base of the Albin Ridge in the west and a collar-shaped occurrence around the Brandberg in the south (Fig. 3B). The sedimentary rocks of the Huab Basin rest unconformably on metamorphozed rocks of the Neoproterozoic Zerrissene and Swakop groups of the Damara Sequence ( Fig. 3B; Nieminski et al., 2018) and also on Cambrian pan-African granites (Goscombe et al., 2017; and references therein). The lowest stratigraphic unit of the Huab Basin is the thin (ca 15 m; Fig. 4) glaciogenic Dwyka Group (Fig. 5), resting directly on metamorphozed pan-African basement rocks (Figs 3B and 6A). The Dwyka succession correlatives (Martin, 1953;Frets, 1969) comprise diamictites, cross-bedded sandstones and conglomerates which occur in small patchy outcrops (Horsthemke, 1992), suggesting deposition under proximal to distal fluvioglacial to fluviolacustrine conditions as outwash fans (Miller, 2008). All of the following stratigraphic units are considered to represent Ecca Group equivalents (Horsthemke, 1992;Ledendecker, 1992). Following up in stratigraphy, the Verbrande Berg Formation (Figs 4 and 5) reaches a maximum thickness of ca 100 m and is largely comprised of carbonaceous and silty shales as well as siltstones (Figs 4 and 6B). In places, a 4 m thick sandstone layer occurs (Miller, 2008). More detailed descriptions of the Verbrande Berg Formation facies are given in Ledendecker (1992) and Holzf€ orster et al. (1999). The siliciclastic deposition most likely took place under cooltemperate post-glacial limnic and low-energy  Pankhurst et al. (2006), Isbell et al. (2008), Pagani & Taboada (2010) and Kent & Muttoni (2020). Please note: The star marks the position of the Huab Basin. fluvial conditions of swamps alternating with shallow lacustrine domains (Horsthemke, 1992). The Verbrande Berg Formation is conformably overlain by the sandstone successions of the Tsarabis Formation (Figs 4 and 6C). The ca 10 to 15 m thick fluvial sediments can be subdivided into a proximal fluvial, a delta and a distal prodelta facies (Ledendecker, 1992;Holzf€ orster et al., 1999) and was deposited within a braided river system gradually changing in a shallow-marine nearshore environment under cold, relatively dry climates (Horsthemke, 1992;Wanke et al., 2000). The Probeer Formation lies conformably on top of the Tsarabis Formation (Fig. 5) and consists of up to 70 m thick massive calcareous aggregates (Fig. 4) of varying colours. The formation thins out rapidly to the west and is only present within the study area as decimetre-sized thin bands (Figs 3B and 6). The origin has been considered by Horsthemke (1992) as largely pedogenic in strongly vegetated soils. Discordantly on top of the Probeer Formation follow the mature, coarse-grained and quartz-rich sandstones of the Gudaus Formation (Figs 4, 5 and 6), with a highly variable thickness of 2 to 30 m (Miller, 2008). The fluvial deposited sandstones were most likely part of a deltaic system with high-energy, littoral sand reworking during a transgression phase (Horsthemke, 1992). This system is characterized by coarsening upward and is responsible for a good sorting and rounding of the material (Miller, 2008). The topmost successions of the Ecca Group of the Huab Basin are the sandstones of the Gai-as Formation (Fig. 4) Schreiber (2002) and Milner (2007). Colours of sample names correspond with their respective formation.

METHODS
Samples were collected in the Huab Basin area and in the Kunene River area in the northern part of Namibia. All Permo-Carboniferous siliciclastic rock samples were representative for the respective facies found in the studied units. Zircon concentrates were separated from 1 to 3 kg whole rock material at the Senckenberg Naturhistorische Sammlungen Dresden (Museum f€ ur Mineralogie und Geologie). After crushing the fresh sample material in a jaw crusher, material was sieved for the fraction from 36 to 400 lm. Heavy mineral separation was achieved from the latter fraction using LST (lithium heteropolytungstate in water) prior to magnetic separation in a Frantz isomagnetic separator. Final selection of the zircon grains for U-Pb dating was carried out by hand-picking under a binocular microscope. When possible, at least 150 zircon grains of all grain sizes and morphological types were selected for each sample. After selection the morphological types according to Pupin (1980), length, width and roundness were determined with a Zeiss EVO 50 scanning electron microscope (Carl Zeiss Microscopy GmbH, Oberkochen, Germany; e.g. G€ artner et al., 2018; and references therein). These characteristics of morphology are aimed to supplement the isotopic data and may help to improve the precision of provenance studies. Zircon size was determined using the software DIPS 2.9 (point electronic GmbH). The degree of zircon roundness was determined following G€ artner et al. (2013). Zircon grains were mounted in resin blocks and polished to half their thickness in order to expose their internal structure (for example, oscillatory growth and older cores). Cathodoluminescence (CL)-imaging was performed using a Zeiss EVO 50 SEM coupled to a HONOLD CL detector operating with a spot size of 550 nm at 20 kV.
The zircon grains were analysed for U, Th and Pb isotopes by LA-SF ICP-MS techniques (laser ablation À sector field inductively coupled plasma À mass spectrometry) at the Museum f€ ur Mineralogie und Geologie (Sektion Geochronologie, Senckenberg Naturhistorische Sammlungen Dresden), using a Thermo-Scientific Element 2 XR sector field ICP-MS (single-collector; Thermo Fisher Scientific, Waltham, MA, USA) coupled to an asi RESOlution 193nm excimer laser (Australian Scientific Fyshwick, ACT, Australia). Each analysis consisted of 15 s background acquisition followed by 30 s data acquisition, using a laser spot-size of 25 to 35 lm and was bracketed by zircon reference material measurements. A common-Pb correction based on the interference-corrected and background-corrected 204 Pb signal and a model Pb composition (Stacey & Kramers, 1975) was carried out if necessary. The necessity of the correction is judged on whether the corrected 207 Pb/ 206 Pb lies outside of the internal errors of the measured ratios (Frei & Gerdes, 2009). Discordant analyses were generally interpreted with care. Raw data were corrected for background signal, common Pb, laser-induced elemental fractionation, instrumental mass discrimination and time-dependent elemental fractionation of Pb/Th and Pb/U using an Excel â spreadsheet program developed by Axel Gerdes (Institute of Geosciences, Johann Wolfgang Goethe-University Frankfurt, Frankfurt am Main, Germany). Reported uncertainties were propagated by quadratic addition of the external reproducibility obtained from the reference zircon GJ-1 (ca 0.6% and 0.5-1.0% for the 207 Pb/ 206 Pb and 206 Pb/ 238 U, respectively) during individual analytical sessions and within-run precision of each analysis. In order to test the accuracy of the measurements and data reduction, the Ple sovice zircon was included as a secondary reference in our analyses, which gave reproducibly ages of ca 337 Ma, fitting with the results of Sl ama et al. (2008). The 207 Pb/ 206 Pb age was taken for interpretation of all zircon grains >1.5 Ga, and the 206 Pb/ 238 U ages for younger grains as recommended by Puetz (2018). For further details on analytical protocol and data processing see Gerdes & Zeh (2006). A UÀPb analysis is concordant when it overlaps within uncertainty with the Concordia. So, it seems to be appropriate to exclude results with a low level of concordance ( 206 Pb/ 238 U age/ 207 Pb/ 206 Pb age 9 100), but very large errors that overlap with the Concordia from interpretation. Thus, an interpretation with respect to the obtained ages was done for all grains within the concordance interval of 90-110% ( 206 Pb/ 238 U age/ 207 Pb/ 206 Pb age 9 100) which is often used (Spencer et al., 2016). In order to exclude lead loss effects, analyses with >2.5% corrected common lead were rejected by default and were not considered for further interpretation (Andersen et al., 2019). Uranium and Pb content and Th/U ratio were calculated relative to the GJ-1 zircon reference and are accurate to approximately 10%. Analytical results of U-Th-Pb isotopes and derived U-Pb ages are given in the supplementary data. The stratigraphic timescale of Gradstein et al. (2012) was used.
In order to calculate the maximum depositional age from detrital zircon grains many methods have been proposed (Coutts et al., 2019;and references therein). Following these studies, the four methods described are considered here: the youngest single grain (YSG; Dickinson & Gehrels, 2009); the youngest grain cluster at 2r (YGC 2r; Dickinson & Gehrels, 2009); the TuffZirc algorithm of the six youngest dates (TuffZirc 6+; Tucker et al., 2013); and the youngest three zircons (YZ3o; Ross et al., 2017). The dated zircon grains used for determining the MDA are in general idiomorphic (low roundness classes) and most likely underwent only few sedimentation cycles (Markwitz & Kirkland, 2018). The maximum depositional age of investigated sedimentary rocks corresponds with inferred depositional timing. Therefore, the obtained ages were considered as good approximations. Hafnium isotope measurements were carried out using a Thermo-Finnigan NEPTUNE multi-collector ICP-MS (Thermo Fisher Scientific) at the Institute of Geosciences, Johann Wolfgang Goethe-University Frankfurt, Frankfurt am Main, Germany, coupled to a RESOlution M50 193 nm ArF Excimer (Resonetics) laser system following the method described in Gerdes & Zeh (2006, 2009. Spots of 40 µm in diameter were drilled with a repetition rate of 4.5 to 5.5 Hz and an energy density of 6 J/cm 2 during 50 s of data acquisition. The instrumental mass bias for Hf isotopes was corrected using an exponential law and a 179 Hf/ 177 Hf value of 0.7325. In the case of Yb isotopes, the mass bias was corrected using the Hf mass bias of the individual integration step multiplied by a daily bHf/bYb offset factor (Gerdes & Zeh, 2009). All data were adjusted relative to the JMC475 of 176 Hf/ 177 Hf ratio = 0.282160 and quoted uncertainties are quadratic additions of the withinrun precision of each analysis and the reproducibility of the JMC475 (2SD = 0.0028%, n = 8). Accuracy and external reproducibility of the method were verified by repeated analyses of reference zircon GJ-1 and Ple sovice, which yielded a 176 Hf/ 177 Hf of 0.282007 AE 0.000026 (2SD, n = 42) and 0.0282469 AE 0.000023 (n = 20), respectively. This is in agreement with previously published results (Gerdes & Zeh, 2006;Sl ama et al., 2008) and with the LA-MC-ICP-MS long-term average of GJ-1 (0.282010 AE 0.000025; n > 800) and Ple sovice (0.282483 AE 0.000025, n > 300) reference zircon at the Frankfurt lab.
The initial 176 Hf/ 177 Hf values are expressed as eHf, which is calculated using a decay constant value of 1.867 9 10 À11 year À1 , CHUR after Bouvier et al. (2008;176 Hf/ 177 Hf CHUR , today = 0.282785 and 176 Lu/ 177 Hf CHUR , today = 0.0336) and the apparent U-Pb ages obtained for the respective domains are shown in the supplementary data A. For the calculation of Hf twostage model ages (T DM ) in billion years, the measured 176 Lu/ 177 Lu of each spot (first stage = age of zircon), a value of 0.0113 for the average continental crust, and a juvenile crust 176 Lu/ 177 Lu NC = 0.0384 and 176 Hf/ 177 Hf NC = 0.283165 (average MORB; Chauvel et al., 2007) were used.
Nonmetric multidimensional scaling (MDS) plots based on the KolmogorovÀSmirnov statistical analysis (Vermeesch, 2013) were produced using the provenance package for the statistic program R 3.6.1 (Vermeesch et al., 2016). MDS produces a point configuration in which similar samples plot close together and dissimilar samples plot far apart in order to compare the sample zircon age data with zircon age data from published studies from southern African structural units (along time) and from other Mesozoic Karoo-aged successions (geographically). Kernel density estimation plots were produced using the detzrcr package for the statistic program R 3.6.1 (Andersen et al., 2018b).
The geochemical analyses of the rock samples were carried out by Actlabs in Ancaster (Ontario, Canada) following the procedure proposed by Panteeva et al. (2003). A lithium metaborate/ tetraborate fusion technique was performed, providing a fast and high quality result. The resulting molten bead is rapidly digested in a weak nitric acid solution, ensuring an entirely dissolved sample. Subsequent analysis was carried out by ICP-OES and ICP-MS.
An overview of the most important results of this study may be found in Table 1. All obtained UÀThÀPb and La-Hf LA-ICP-MS measurements including morphological features of each grain can be found in the supplementary data.

RESULTS
The sampled Dwyka-aged sandstone of the Kunene section was massive to poorly bedded, tan-coloured, highly calcareous and almost pebble-free. It contained a high number of trains, lenses and irregular layers of grit and conglomerate. The underlying laminated shales carried widely scattered pebbles and boulders. The sampled layered diamictites of the Dwyka Group of the Huab Basin contained subangular to rounded clasts of basement rock types. Clasts as well as the pale green greyish, sandy to argillaceous matrix are subangular to subrounded. In places layers of thin matrixsupported conglomerates and fine to mediumgrained sandstones occurred. The sandstones showed thin conglomeratic interbeds and planar to trough cross-bedding. The Verbrande Berg Formation sample was a yellow brownish feldspathic siltstone to mudstone with haematitic concretions. The Tasarabis Formation sample is represented by a whitish to whitish-grey, coarsegrained, highly feldspathic, poorly sorted sandstone. The overlaying Gudaus Formation sample is represented by a coarse-grained to gritty,  Herron (1988). (C) Ternary diagram of the chemical index of alteration (CIA) after Nesbitt & Young (1982) and Nesbitt & Young (1984), including the correction trend for K metasomatism according to Fedo et al. (1995). Numbered symbols indicate typical values of reference lithologies: 1-gabbro; 2-tonalite; 3-granodiorite; 4-granite; 5-A-type granite; 6-charnokite; 7-potassic granite. Ideal weathering trends of upper continental crust-type source lithologies would be parallel to the predicted weathering trend (Nesbitt & Young, 1984). The statistically modelled t-index weathering trend (von Eynatten, 2004) is based on data obtained from the world's major rivers and areas under erosion (McLennan, 1993). (D) La/Th versus Hf diagram after Floyd & Leveridge (1987) and Gu et al. (2002) to determine average source composition for the investigated sediments. LCC: Lower Continental Crust; UCC: Upper Continental Crust. All diagrams are based on data given in supplementary material.
quartz-rich, porous, bedded very well-sorted sandstone of brownish colour. Quartz grains are normally well to very well-rounded. The topmost Gai-as Formation K-felspar-rich sandstone was greyish and fine-grained. Micritic calcite, dolomite and ankerite were present. Whole rock geochemical data has been obtained for nine samples of early Permian sandstones and shales of the northern Namibian Huab Basin (Fig. 7) in order to determine recycling, weathering and average composition of potential source areas. The sandstones and shales associated with the northern Namibian Karoo-aged successions have SiO 2 values mostly between 65 and 92 wt% (supplementary data) and correspond to quartz-rich sandstones of Crook (1974). Just sample Nam 399 yields SiO 2 = 59 wt%, which is classified as quartzintermediate sandstones. Chondritic normalized rare earth element (REE) diagrams of all analysed samples show a negative Eu-anomaly typical for plagioclase-rich rocks, as partially observed within the petrographic description (Fig. 7A). Log (SiO 2 /Al 2 O 3 ) versus log (Fe 2 O 3 / K 2 O) diagram after Herron (1988) classifies the investigated samples as Fe-sands and Fe-shales, respectively (Fig. 7B), as already suggested for the Verbrande Berg sample Nam 397 by its ferruginous cement. The chemical index of alteration is a tool to gauge the effects of weathering and alteration of the composition of siliciclastic sedimentary rocks. In the CN-A-K plot, the sedimentary samples follow a relatively steep trend along the CaO+Na 2 O-Al 2 O 3 conode (Fig. 7C), following the natural weathering trends (Nesbitt & Young, 1984;von Eynatten, 2004) and the post-Archean average Australian shale composition (PAAS; Taylor & McLennan, 1985). On a La/Th versus Hf diagram, an upper continental crust (UCC) provenance composition is shown by La/ Th ratios between 0.80 and 3.92. Hf concentrations are typical of felsic to felsic/basic source, with no sedimentary recycling trend (Fig. 7D).
This study analysed detrital zircon grains from nine Permo-Carboniferous sandstone samples of the Dwyka Group of the Kunene area as well as the Dwyka and Ecca groups (Verbrande Berg, Tsarabis, Gudaus and Gai-as formations) of northern Namibia (Fig. 4). In total, 1441 zircon grains were investigated with respect to their morphological features, including width, length and roundness value after G€ artner et al. age group in the range of ca 1.80 to 2.00 Ga (30% of all concordant analyses) (Fig. 10). Archean ages are scarce but could be observed within all investigated samples except sample Nam 402 (3% of all concordant analyses) (Fig. 10).   Group) of northern Namibian Karoo-aged sediments (Fig. 11). The detrital zircon Hf composition yields information concerning the composition and origin of the magma from which the zircon will have crystallized (juvenile, evolved or mixed). All investigated samples, besides Nam 389 and Nam 391 show a wide spread in eHf values, ranging from À31.3 to 12.1. A vast majority of analyses of ca 67% fall into the field of an evolved crustal composition, In the following, there is a more detailed sample description, sorted after the respective section and after that from bottom to the top of the successions. A stratigraphic classification of the samples is given in Fig. 6. Nam 399, S21°05'49.4", E14°09'54.0", diamictite, uppermost Carboniferous(?), Dwyka Group The reddish-grey diamictite was collected about 28 km west of the Brandberg Massif in a gully lying directly above the granitic pre-Karoo basement (Figs 3B and 6A). The 151 retrieved zircons yield mean lengths and widths of 108 µm and 57 µm, respectively (Fig. 8). Of them, ca 90% show a poor to a very good degree of roundness (classes 4 to 8; Fig. 9). The 131 UÀThÀPb analyses gave 83 concordant ages, ranging from 476 AE 8 to 2976 AE 42 Ma (Fig. 10). One prominent main age group ranges from ca 480 to ca 550 Ma (42% of all near concordant analyses) and has one peak at ca 515 Ma. Minor age groups are from ca 900 to ca 1200 Ma (23% of all near concordant analyses) and from ca 1800 to ca 2000 Ma (10% of all near concordant analyses) (Fig. 10).

Dwyka Group
Sample Nam 401 is a bedded, yellowish to brownish sandstone, which was collected in the same gully as sample Nam 399. The 148 separated zircons inherit mean lengths and widths of 151 µm and 76 µm, respectively. About 93% are very poorly to well-rounded (classes 3 to 7; Fig. 9). The 148 performed UÀThÀPb analyses gave 75 concordant ages, ranging from 471 AE 9 to 2634 AE 10 Ma. The detrital zircon UÀPb age distribution is very similar to sample Nam 399, with one major age group ranging from ca 480 to ca 530 Ma (47% of all near concordant analyses) peaking at ca 505 Ma, and minor Mesoproterozoic to early Neoproterozoic and Palaeoproterozoic detrital zircon age clusters ranging from ca 900 to ca 1200 Ma (15% of all near concordant analyses) and from ca 1800 Ma to ca 2200 Ma (17% of all near concordant analyses), respectively (Fig. 10).
Nam 389, S21°06'10.9", E14°18'59.6", sandstone, Lower Permian, Ecca Group, Tsarabis Formation The massive pinkish sandstone was collected along the road D2342, approximately 10 km west of the Brandberg Massif (Figs 3B and 6C). One hundred fifty one zircon grains were investigated and have mean lengths and widths of 110 µm and 59 µm, respectively (Fig. 8). About 84% of the grains are almost completely unrounded to rounded (classes 2 to 6; Fig. 9). Of 152 UÀThÀPb analysis, 144 yield concordant ages between 288 AE 5 to 3085 AE 15 Ma. One main age group from ca 1800 to ca 2000 Ma can be determined (69% of all concordant analysis), peaking at ca 1850 Ma. Minor age groups are from ca 500 to ca 650 Ma (12% of all near concordant analyses) and from ca 950 to ca 1100 Ma (8% of all near concordant analyses; Fig. 10).
Nam 391, S21°06'11.0", E14°18'59.6", sandstone, Gudaus Formation, Lower Permian The massive brownish sandstone of the Gudaus Formation was collected directly above sample Nam 389 (Figs 3B and 6C). One hundred and forty nine grains were investigated with respect to their mean length and width, which resulted in 159 µm and 86 µm, respectively (Fig. 8). About 70% of the zircons are almost completely unrounded to fairly rounded (classes 2 to 6; Fig. 9). The 153 UÀThÀPb analyses gave 134 concordant ages, ranging from 447 AE 8 to 2914 AE 16 Ma. One main age group appears from ca 1800 to ca 2000 Ma (57% of all near concordant analyses) peaking at ca 1850 Ma. Minor age groups occur between ca 490 to ca 650 Ma (13% of all near concordant analyses) and between ca 950 to ca 1100 Ma (14% of all near concordant analyses) (Fig. 10).

Timing of deposition of the Huab Basin strata
The use of maximum depositional ages (MDAs) in detrital zircon studies is still a valid method and has been used in many recent publications (Bahlburg et al., 2020;Barrett et al., 2020;Sharman & Malkowski, 2020). Nevertheless, the best method for estimating MDAs is under strong debate (Dickinson & Gehrels, 2009;Sharman et al., 2018;Coutts et al., 2019;and references therein).
As for the small amounts of Permian UÀPb zircon ages obtained from the studied samples (Fig. 13), only samples Nam 389 and Nam 397 yield somewhat reliable MDAs. The youngest single grain (YSG) method of sample Nam 397 (Verbrande Berg Formation) yields an age of 290.0 AE 5.4 Ma (Fig. 12). The youngest three dated zircon grains (overlapping in error) method (Y3Zo) resulted in a slightly older age of 292.2 AE 9.2 Ma (Fig. 12). The oldest ages were achieved by applying the TuffZirc 6+ method and by the youngest single population (YSP) method resulting in MDAs for the Verbrande Berg Formation of 294.7 + 3.2/À5.0 Ma and 294.5 AE 6.7 Ma, respectively (Fig. 12). As Herriott et al. (2019) identified the YSP method as the most robust way of estimating MDAs using LA-ICP-MS UÀPb ages, the age of 294.5 AE 6.7 Ma (Asselian to Sakmarian) will be considered here as best age approximation. The obtained age hints towards rapid deposition after the last Dwyka deglaciation cycle, which probably ended in the Aranos/Karasburg Basin area at ca 295 Ma (Zieger et al., 2020b). Furthermore, the Verbrande Berg Formation may be correlated with the lower parts of the Prince Albert Formation of the Main Karoo Basin (South Africa) and the Karasburg Basin (Namibia; Fig. 5). Equivalent successions of the Aranos Basin may be the Nossob and the lower Mukorob formations. (Fig. 5).
Only one zircon grain of sample Nam 389 of the Tsarabis Formation yielded an early Permian age of deposition, which lies within the age range for the lower Huab Basin successions proposed by Miller (2008). Therefore, the YSG method provided an age of 288.4 AE 4.8 Ma (Sakmarian-Artinskian; Fig. 12). This age correlates the Tsarabis Formation with the middle parts of the Mukorob Formation (Aranos Basin), the Prince Albert Formation of the Karasburg and Main Karoo basins as well as the Rio Bonito Formation of the Paran a Basin (Fig. 5).
As the MDAs of northern and southern Namibian Karoo-aged basins overlap in error (Zieger et al., 2019; Fig. 13), they imply a contemporaneous sedimentation of the Karoo Supergroup successions of Namibia. Thus, this study proposes a very small time interval between the evolution of both Namibian parts of the Karoo I rift system (Frizon de Lamotte et al., 2015; Fig. 1). It should be noted that to our knowledge, no Permian zircon producing sources are known from the vicinity of the study area. Therefore, the deduced MDAs suggest a rapid mixing of already eroded basement material with small amounts of ash fall tuffs, as zircon grains of Permian age are in general considered to be derived from ash fall deposits (Rubidge et al., 2013;McKay et al., 2015;Viglietti et al., 2018). Possible source areas for these zircon grains are West Antarctica, the Patagonian terrane and the Central Andes (Stollhofen et al., 2000a;Viglietti et al., 2018).
Despite their popularity, Herriott et al. (2019) revealed flaws estimating MDAs of sedimentary successions younger than ca 250 Ma using LA-ICP-MS UÀPb zircon ages. Therefore, imprecise MDA measurements of Palaeozoic sedimentary successions cannot be ruled out. Furthermore, polycyclicity of the grains is also possible due to transport within rock fragments derived from nearby sources acting as an intermediate sediment repository supplying recycled material.

Implications from detrital zircon grain morphology and isotope analysis
All investigated samples yield Cambrian to late Neoproterozoic (500 to 650 Ma; >25% of the near concordant age determinations), Mesoproterozoic (1000 to 1200 Ma; >20% of the near concordant age determinations) and Palaeoproterozoic (1800 to 2050 Ma; >18% of the near concordant age determinations) detrital zircon age groups of changing portions (Fig. 10). The most proximal protosources of Palaeoproterozoic aged zircon may be the Epupa Metamorphic Complex located at the southern edge of the Angolan Shield (Kr€ oner et al., 2010(Kr€ oner et al., , 2015Lehmann et al., 2020) and the Kamanjab Inlier situated within the northern Damara orogen (Kleinhanns et al., 2015). Further distant Palaeoproterozoic zircon producing protosources may be the Richtersveld Magmatic Arc and the Orange River Group, which are part of the Namaqua Sector (Van Niekerk, 2006;Pettersson et al., 2007;Minnaar, 2011;Hofmann et al., 2014;Macey et al., 2017). Remarkably, the majority of Palaeoproterozoic zircon grains with ages of ca 1.8 to 1.9 Ga show a narrow range of Ɛ Hf values ranging from ca À3 to 3, suggesting a single zircon producing crustal source (Fig. 11) and no sedimentary mixing.
Late Mesoproterozoic sedimentary protosources are scarce in the northern part of Namibia but have been reported within the Epupa Metamorphic Complex (Kr€ oner & Rojas-Agramonte, 2017). More distant protosources are the granitoids of the Namaqua Sector located in the southern part of Namibia and in South Africa (Raith et al., 2003;Clifford et al., 2004;Macey et al., 2018) and the Irumide Belt of eastern Zambia (De Waele et al., 2009). The portion of late Mesoproterozoic aged UÀPb zircon grains decreases upsection the Huab Basin stratigraphy from 18% comprised by the lower diamictite of the Dwyka Group (sample Nam 399) to 5% by a sandstone of the Gai-as Formation (sample Nam 394; Ecca Group; Fig. 10).
The studied samples yield high amounts of late Neoproterozoic to early Cambrian detrital UÀPb zircon ages (Fig. 10), probably caused by an abundance of pan-African zircon UÀPb age   (4) group (1) group (3) group (2) group (3) group (1) group (2) group (4) bearing protosources situated in the vicinity of the study area (Fig. 3A), namely the Damara Belt as well as the Copperbelt. Cambrian granitoids of the Kaoko Belt directly underlie the successions of the Huab Basin (Goscombe et al., 2017; and references therein; Fig. 3A). The occurrence of pan-African aged zircon grains decreases significantly with the onset of the sedimentation of the Tsarabis Formation successions, where only 12% of the detrital zircon record is comprised by the latter age fraction (Fig. 10). The shift in zircon age pattern may be explained by a decrease of the Huab Basin catchment area (see further discussion). Using traditional sedimentological methods, answering provenance questions becomes increasingly difficult given composite sedimentary features occurring during a major glaciation phase, especially in a smaller scale (Le Heron et al., 2019). The samples of the Kunene section are part of a north-western Namibian system of most likely ice-carved Permo-Carboniferous valleys (Martin, 1981). These are confined by complex patterns of striation orientations, crossbedding, and boulder fabrics that lead to a very confusing picture (Du Toit, 1921;Martin & Wilczewski, 1970;Rosa et al., 2019). Previous publications showed that the use of detrital zircon offers great opportunity when interpreting Dwyka glaciation ice sheet dynamics (Griffis et al., 2019;Zieger et al., 2019).
Based on their distinct detrital zircon UÀPb age patterns the investigated samples may be subdivided into four groups: (1) Dwyka Group sediments of the Kunene section (samples Nam 478 and Nam 479) with prominent lower Mesoproterozoic and Palaeoproterozoic detrital zircon age clusters; (2) Dwyka Group sediments of the Huab Basin with a lower Cambrian detrital zircon age peak (Nam 399, Nam 401 and Nam 402); (3) the Ecca Group sediments of the Verbrande Berg Formation (Nam 397), which are similar to group (2) samples but yield unique Permian detrital UÀPb zircon ages; and (4) the youngest successions of the Huab Basin, namely the Tsarabis, Gudaus and Gai-as formations with one prominent Palaeoproterozoic detrital zircon age peak (Nam 389, Nam 391 and Nam 394).
The visual inspection of detrital zircon age spectra is highly subjective. This issue can be solved by using statistical analysis revealing their similarities and differences (Saylor et al., 2013;Bahlburg & Berndt, 2016;Vermeesch et al., 2016;Sundell & Saylor, 2017;Huber et al., 2018;Bahlburg et al., 2020). For that reason, the application of MDS offers a great opportunity in obtaining a relative 'sameness' between the investigated samples (Vermeesch et al., 2016;Sundell & Saylor, 2017). In doing so, the authors tried to compare the age spectra of our obtained samples with published age spectra of southern African basement units as well as Permo-Carboniferous sedimentary successions.
Although resting on Mesoproterozoic to Palaeoproterozoic basement (Fig. 3), detrital zircon grains extracted from group (1) samples of the Kunene area yield 22 to 25% Cambrian to Lower Ediacaran (ca 500-600 Ma; pan-African) ages (Fig. 10). Interestingly, the sediments of the Kunene section yield 5 to 7% Archean ages in the range from ca 2.5 to 3.0 Ga within their detrital zircon age record, respectively (Fig. 10), which is high in comparison to other Namibian Karoo-aged sedimentary successions ranging between 1 to 2% (Zieger et al., 2019(Zieger et al., , 2020a. Given the detrital zircon age pattern of group (1) samples, the application of KolmogorovÀ Smirnov MDS revealed a highest statistical dissimilarity to UÀPb zircon record of the Palaeoproterozoic terranes and shields (Fig. 13A). Smaller dissimilarities could be observed with Mesoproterozoic and Neoproterozoic terranes. The closest statistical neighbours to group (1) samples are the Neoproterozoic Mozambique and Kaoko belts (Fig. 13A) but still there is no clear attribution to any structural unit possible. Thus, several crustal sources may have contributed towards the detrital zircon record of group (1) samples. In order to compare Hf isotopic signatures of the studied samples with southern African basement units the ratio between UÀPb age and Ɛ Hf value was used (Fig. 13B). The latter approach reveals the Mesoproterozoic Irumide Belt and Barue Complex as well as the Neoproterozoic Mozambique Belt as closest statistical neighbours to group (1) samples (Fig. 13B), maybe pointing towards sediment contribution from these three basement complexes.
Zircon grain morphology is increasingly used in detrital zircon provenance studies (G€ artner et al., 2013;Zoleikhaei et al., 2016;Shaanan & Rosenbaum, 2018;Zieger et al., 2020a;Zeh & Cabral, 2021). Because of their hardness, the crystals are being rounded very slowly within fluvial and littoral regimes (Garzanti et al., 2015). Therefore, well-rounded zircon grains may be evidence for ultra-long distance transport. Although it is assumed here that the vast majority of the investigated grains became rounded due to transport, other processes may also initiate grain roundness. Physicochemical alteration (Mager, 1981;Deer et al., 2001;Tichomirowa et al., 2005), erosion effects of transporting magma (G€ artner et al., 2016, and references therein) as well as inherited rounded grains in granites (Tichomirowa et al., 2001;Roger et al., 2004) are also possible sources of rounded zircon grains. Due to the limited occurrence of such grains (Tichomirowa et al., 2001;Tichomirowa, 2001;Tichomirowa, 2002;Hoskin & Black, 2002), only zircon roundness will be considered as a result of sedimentary transport and recycling.
The zircon grains included within group (1) samples are generally fairly rounded (Fig. 9), consequently pointing towards repeated and/or long distance transport of the sediments. All Archean grains yield roundness values ≥5 (Fig. 9) ruling them out as 'source-to-sink' material directly derived from the southern Congo Craton margin (Seth et al., 1998). It is more likely that zircon grains from the latter area became eroded and underwent repeated sedimentary recycling processes, and were eventually deposited as Dwyka Group sediments of the Kunene section. Width and length values of the zircon grains mainly plot within a narrow range (Fig. 8), maybe pointing towards sorting effects during sedimentation of the Kunene section successions. Interestingly, the zircon grain width versus length dispersal pattern does not differ significantly, although the sediments change upsection from shaly diamictites towards finegrained sandstones. The latter similarities in zircon morphology and detrital zircon age patterns are interpreted to constrain a similar sediment transport history of the Kunene section successions. The variety of UÀPb detrital zircon age groups obtained within samples Nam 478 and Nam 479 (Fig. 10) may be evidence of a homogenized sedimentary cover of Gondwana persisting during the Dwyka period. This follows the interpretation of Andersen et al. (2016) for the sedimentary rocks of the Main Karoo Basin and confirms the findings of Zieger et al. (2019) for the Dwyka Group sedimentary rocks of the Aranos Basin located further south. Thus, the detrital zircon pattern of group (1) samples may not inherit any specific source information, as the original source patterns were most likely obscured by several cycles of sedimentary transport, storage and erosion.
Sample groups (2) and (3) yield very similar detrital zircon age patterns (Fig. 10). The sampled lower sections of the Huab Basin [groups (2) and (3)] are represented by the diamictites and sandstones of the Dwyka Group sedimentary rocks [group (2) and shales of the Verbrande Berg Formation of the lowermost Ecca Group (group (3)], resting directly on pan-African Damara granites (Figs 3 and 6A). The detrital UÀPb zircon age record of both groups yield high portions of Cambrian to late Neoproterozoic ages (49-55%) and low portions of Rodiniarelated upper Mesoproterozoic (14-18%) as well as mid Palaeoproterozoic ages (7-14%). Thus, the detrital zircon age dispersal pattern differs from group (1) samples (Fig. 10). Applying Kol-mogorovÀSmirnov based MDS to group (1) and (2) samples further suggests a similar recycling history of the samples, as they plot very close to each other (Fig. 13A). The closest statistical neighbouring basement unit is the Saldania Belt of the Neoproterozoic terrane realm (Fig. 13A). The remaining southern African basement complexes plot very far away within the MDS range. Differences are further suggested by slightly negative to positive Ɛ Hf values of group (2) samples, pointing towards an evolved to moderately juvenile magma composition, whereas Ɛ Hf values of group (3) samples mainly fall into the slightly juvenile to juvenile fields (Fig. 11). This difference becomes visible comparing age/ Ɛ Hf using KolmogorovÀSmirnoff MDS (Fig. 13B): The closest statistical neighbour of group (2) samples is the far east located Mozambique Belt, whereas group (3) is confined to the proximate Damara Belt. These findings may point towards a source area of Dwyka Group sediment samples [group (2)] from the east. This is in line with a proposed sediment flux from the east during the Permo-Carboniferous glaciation (Visser, 1987;Griffis et al., 2019;Zieger et al., 2019).
The increased occurrence of Permian aged detrital zircon grains within Karoo-aged shales observed by Zieger et al. (2020b) could also be detected within the shales of the Verbrande Berg Formation (Fig. 10). As shale deposition takes place further away from the continent, the sedimentary input of the latter fades and the input of airborne ash-fall tuffs may be much more pronounced.
Despite their similar detrital zircon UÀPb age patterns the morphology of the investigated zircon grains of groups (2) and (3) differs. Mean roundness values of group (2) sample zircon grains range from 5.49 and 5.81, whereas zircon grains of group (3) yield a mean roundness of 4.27 (Table 1). Deduced from their zircon roundness values, group (2) samples most likely underwent moderate recycling obscuring their protosource, comparable to samples of group (1). This goes well with a proposed source area maybe located within the Mozambique Belt about 2500 km away from the final deposition location (see discussion above). In contrast, the poorly rounded grains of the Verbrande Berg Formation may be witness for a more local source area of the zircon grains, which is in line with the earlier proposed source located in proximity to the Damara Belt.
As the length versus width dispersal pattern can be used in distinguishing between sorting effects (G€ artner et al., 2017), length and width values are also proxy for the environment prevailing during sedimentation (Allen, 1971;McLaren & Bowles, 1985;Watson et al., 2013). This relation becomes apparent as the Verbrande Berg Formation was deposited under low-energy swampy to limnic conditions (Horsthemke, 1992), yielding the smallest mean zircon grain sizes of all investigated samples (length = 87 µm, width = 42 µm; Table 1). Samples Nam 401 (sandstone) and Nam 402 (diamictite) of group (2) are virtually indistinguishable from one another yielding a wide range of length and width values, which are much more narrow than sample Nam 399 (diamictite; Fig. 8). These findings may point towards increasing transport energies during deposition.
Among the samples investigated for this study, group (4) samples (upper Ecca Group) stand out completely as they yield major portions of upper Palaeoproterozoic detrital zircon ages (58-77%; Fig. 13). The somewhat unique detrital zircon isotopic composition is also suggested by zircon grain roundness values of 4.83 to 3.62 (fairly rounded to very poorly rounded; Table 1), which differ significantly from the older Karoo-aged successions of the Huab Basin and Kunene sections (Table 1; Fig. 9). The low roundness values are interpreted as a possible indicator of a proximal source area of the sediments. Mentioned features of group (4) samples let them stand out among other southern Gondwanan Karoo-aged successions (Fig. 13), as studies depicting local sedimentary sources (Veevers & Saeed, 2007) with reported mainly Palaeoproterozoic to Archean detrital zircon age distributions are scarce (Veevers & Saeed, 2007;Jansson, 2010). The reasons for basement derived detrital zircon age patterns remain unknown but involve the presence of exposed basement rocks as well as short transport distances impeding sedimentary mixing and rounding of the zircon grains.
These studies are in great contrast to the vast majority of Karoo-aged succession detrital zircon age records (Andersen et al., 2016;Canile et al., 2016;Zieger et al., 2019). Locally derived sediments and short transport distances contradict the model proposed by Andersen et al. (2016) for the Main Karoo Basin that was applied by Zieger et al. (2019) to the Namibian Karoo-aged basins. The closest Palaeoproterozoic detrital zircon producing basement unit is the Kamanjab Inlier located in the north-east, representing the most likely source of the sediments (Figs 13A  and 14).
The detrital zircon provenance data review by Andersen et al. (2016) revealed a not yet recognized sedimentary recycling pattern and questions how such data may be interpreted. Since then an increasing interest concerning the behaviour of detrital zircon is apparent (Pereira et al., 2016;Andersen et al., 2018;Andersen et al., 2019;Zieger et al., 2019;Andersen et al., 2020;Zieger et al., 2020a,b;Pereira & Gama, 2021) but fluctuations and their drivers within such a system have been overlooked completely. Changes in the detrital zircon age and morphology patterns were detected within the Huab Basin successions (Figs 9, 10 and 13). During the course of its development, the Huab Bain strata underwent a severe change of climatic conditions from cold to temperate climates in the early Permian with a high water supply and a subsequent shift towards warmer, drier periods in the middle Permian (Horsthemke, 1992). The cold to temperate climates coincide with the deposition of early Permian aged sample groups (2) and (3). As the high water supply entails vast river systems with large catchment areas (Miller, 2008; Fig. 14A), restoring already mixed Gondwana peneplain material (Andersen et al., 2016). This model may explain the diversity provided by the detrital zircon record of both latter sample groups. In contrast, the drastic change in detrital zircon pattern of sample group (4) coincides with the shift towards a lesser water supply. This climate shift is suggested by the occurrence of calcretes within the lower Tasarabis Formation (Retallack, 1983;Cadle et al., 1993). Consequently, the climate shift resulted in significantly reduced catchment areas with perennial river systems ( Fig. 14B; Horsthemke , 1992). These smaller catchment areas may be confirmed by a significantly reduced variety of detrital zircon age clusters and degree of grain roundness of group (4) samples (Fig. 10), eroding only the Palaeoproterozoic basement represented by the Kamanjab Inlier (Kleinhanns et al., 2015;Fig. 14B) explaining the sudden shift of the detrital zircon age signal. Nevertheless, a change in the tectonic regime cannot be excluded causing the detrital zircon age shift. Due to a low temporal resolution of the late Palaeozoic southern African rifting history being available more research is needed (Frizon de Lamotte et al., 2015; and references therein).

The northern Namibian Karoo-aged basins within the southern Gondwanan framework
The obtained isotopic compositions of the investigated detrital zircon grains of this study were compared with the southern Gondwanan detrital zircon record in order to reveal resemblances between the widespread occurrences of Karoo-aged strata. This was achieved by comparing late Palaeozoic southern Gondwanan UÀPb age distributions utilizing a Kol-mogorovÀSmirnov MDS plot (Fig. 13C). For clarification purposes, the formations of the Karoo-aged basins were classified and colourcoded with available detrital zircon age records after the respective basin (Fig. 13C). The application resulted in a division of the investigated samples into three distinct clusters within the MDS space: The first cluster represents sample group (1) consisting of samples Nam 478 and Nam 479 (Fig. 13C). The largest dissimilarities with samples of latter group are apparent with the two side clusters of the Main Karoo Basin and the Aranos + Karasburg Basin realms ( Fig 13C). Smaller dissimilarities can be observed with the main Aranos + Karasburg Basin realms. The smallest dissimilarities are to the Rio do Sul Formation of the left branch of the Paran a Basin realm (Fig. 13C), maybe suggesting a close evolutionary connection between both glacial induced successions (Castro et al., 2000). The second distinct cluster represents sample groups (2) and (3) and shows small statistical dissimilarities to the main Aranos + Karasburg, Paran a, Main Karoo and Tepuel Basin realms (Fig. 13C). The third cluster consisting of samples Nam 389, Nam 391 and Nam 394 coincides with sample group (4) (Fig. 13C). It becomes apparent that sample group (4) has high statistical dissimilarities with all investigated detrital zircon distributions of Permo-Carboniferous southern Gondwanan Karoo-aged successions. As they plot separately, they illustrate the unique character of the detrital zircon record of the upper part of the Huab Basin successions, namely the Tsarabis, Gudaus and Gai-as formations. The smallest statistical dissimilarity is with the Lukuga Formation of the Congo Basin yielding also high portions of upper Palaeoproterozoic  Horsthemke (1992). See text for further explanations. detrital zircon UÀPb ages. As the occurrence of the Lukuga Formation is located ca 2500 km to the north (Linol et al., 2016), the proposed similarities may be coincidental. Andersen et al. (2016) proposed a sedimentary recycling and homogenization system for today's South Africa (Main Karoo Basin), which was extended by Zieger et al. (2020b) for the southern Namibian region. This system existed at least until the Cretaceous (Andersen et al., 2020;Zieger et al., 2020a). The findings of this study support the implications drawn from southern Namibian detrital zircon patterns published by Zieger et al. (2020b) and assign them towards northern Namibia (Fig. 14A). In addition, the comparison of sample groups (1) to (3) with the detrital zircon record of late Palaeozoic successions presented within the MDS plot reveals small statistical dissimilarities (Fig. 13C). Therefore, these patterns may translate into similar detrital zircon age populations across southwestern Gondwana, assuming a sedimentary homogenization system. In contrast, sample group (4) of this study does not fit in the recycling scheme of Andersen et al. (2016). The latter sample group shows high statistical dissimilarities with other southern Gondwana detrital zircon age patterns (Fig. 13C) and, therefore, may represent a unique sedimentary regime or process not yet found within the late Palaeozoic Gondwanan sedimentary record (see discussion above).

CONCLUSIONS
This study investigated 1430 detrital zircon grains of the northern Namibian Karoo-aged basins with respect to their grain morphology and their isotopic composition.
The maximum depositional ages obtained from the Verbrande Berg and from the Tsarabis formations are at 294.5 AE°6.7 Ma (Asselian-Sakmarian) and 288.4 AE 4.8 Ma (Sakmarian-Artinskian), respectively. Thus, the sedimentary successions of the lower Huab Basin may be correlated with the Prince Albert Formation of the Main Karoo and Aranos basins, suggesting a contemporaneous evolution of the northern and southern branch of the Karoo I rift system in the southern African region.
Due to fluvial sedimentary conditions constraining the evolution of the northern Namibian Permo-Carboniferous successions, they occupy a special position among the Karoo Supergroup framework, by means of their varying detrital zircon record and importance of depositional timing for understanding the Karoo system evolution. Inferred by their detrital zircon isotopic composition, the nine investigated siliciclastic rock samples can be subdivided into four different groups: The stratigraphically lowest three groups yielded a wide variety of Permian to Palaeoproterozoic detrital UÀPb zircon ages in changing portions, suggesting sedimentary sources dispersed over a vast catchment area, which became mixed during transport fitting well with a proposed homogenization of Gondwana cover sequences obscuring their original source. During the proposed ultra-long transport, the zircon grains underwent significant physical abrasion and rounding. Our approach questions the application of 'sourceto-sink' models within supercontinent environments proposed by many authors when interpreting detrital UÀPb data. In fact, the suggested method uses the interplay between detrital zircon morphology and isotopic data to constrain recycling paths within the Gondwanan sedimentary framework.
The topmost three formations of the Huab Basin yielded almost exclusively unrounded zircon grains of Palaeoproterozoic age. Therefore, the sedimentary rocks were most likely derived from the proximal Kamanjab Inlier located in the north-east and induced by short transport distances. The input of local basement sources has been interpreted as a decrease in catchment area and, eventually, shorter transport distances, possibly induced by a shift towards drier climatic conditions. Thus, this study showed that the use of detrital zircons based on their UÀPb ages and grain morphometry may be a valid but mostly unused tool offering great opportunities for reconstructing palaeoclimate conditions. access funding enabled and organized by Pro-jektDEAL.

DATA AVAILABILITY STATEMENT
The data that supports the findings of this study are available in the supplementary material of this article.