3.1 Introduction
Tectonic and climatic changes during the Cenozoic (66 Ma to present), both on global and regional (southern Africa) scales, have profoundly affected land surface processes, the development of different landforms, and the extent to which such signatures are preserved or destroyed in the landscape. An understanding of these issues frames discussion of landscape evolution in southern Africa during the Cenozoic, and thus the reasons why today’s landscapes exist. As such, Cenozoic landscape evolution can also inform on the evolution of land ecosystems (Bamford et al., this volume), large herbivores (Brink, this volume), and hominids (Heaton, this volume; Carlson and Edlund, this volume).
The southern African landscape provides a good opportunity to examine the interplay between different drivers of landscape change, including tectonics and climate (e.g. Partridge and Maud, Reference Partridge and Maud1987, Reference Partridge, Maud, Partridge and Maud2000; Botha and de Wit, Reference Botha and de Wit1996; Séranne and Anka, Reference Séranne and Anka2005; Moore et al., Reference Moore, Blenkinsop and Cotterill2009; Cotterill and de Wit, Reference Cotterill and de Wit2011; Kounov et al., Reference Kounov, Viola, Dunkl and Frimmel2013). Different studies have taken various viewpoints as to which driver is most dominant (e.g. Champagnac et al., Reference Champagnac, Valla and Herman2014), but there are two observations to make based on these viewpoints. First, it is likely that the relative importance of these drivers varied over time and space, and thus that one viewpoint is unlikely to fully describe the evolving nature of the land surface over geological time, or adequately explain the way a given landscape appears today. Second, there are feedback loops between tectonic and climatic processes that (i) are important to consider in evaluating their interrelationships, and (ii) are little understood in southern Africa compared to other areas worldwide. These issues are discussed in detail in Section 3.5 with respect to developing an understanding of landscape evolution in southern Africa during the Cenozoic.
This chapter discusses the tectonic and climatic contexts of landscape evolution in southern Africa on regional to continental spatial scales and over Cenozoic timescales. As such, it is complementary to the shorter time frame discussed by Knight and Grab (this volume).
3.2 Africa’s tectonic and geologic setting
Africa as a whole represents an intracratonic setting, as evidenced by the relatively stable and ancient land surface and its old Precambrian basement of the interior. The tectonic history of southern Africa has been recently summarised by Watkeys (Reference Watkeys, Johnson, Anhaeusser and Thomas2006). This history focuses on the breakup of Gondwana during the period ~180–90 Ma, but has wider implications for changes in continental-scale climate and land surface processes as a result of continental drift and development of ocean circulation patterns in emerging ocean basins (Reyment and Dingle, Reference Reyment and Dingle1987). The continental interior of southern Africa, comprising the Kaapvaal craton, was not strongly affected by tectonic deformation at this time, although sheared and mountain-building zones including the Lebombo, Cape Fold and Gariep belts were formed around craton peripheries prior to this period (James et al., Reference James, Fouch, VanDecar and van der Lee2001). Tectonic subsidence may have contributed to the formation of accommodation space on the Kalahari and Congo craton surface, which was overlain by volcanic and sedimentary sequences such as of the Witwatersrand, Pongola, Ventersdorp, Transvaal, Olifantshoek and Karoo Supergroups which were formed during the Mesozoic and Cenozoic (Bumby and Guiraud, Reference Bumby and Guiraud2005; Said et al., Reference Said, Moder, Clark and Ghorbal2015).
Over long (106–108 year) timescales, the interplay of several factors has contributed to the evolution of the African land surface (Ashwal and Burke, Reference Ashwal and Burke1989). Climatic changes during the Cenozoic and earlier represent one such factor (outlined in Knight and Grab, this volume), but climatic feedbacks arising from tectonic and volcanic factors are also significant (e.g. Sepulchre et al., Reference Sepulchre, Ramstein, Fluteau, Schuster, Tiercelin and Brunet2006). Mantle swells and continental rifting are important tectonic and volcanic controls on the African land surface (Ashwal and Burke, Reference Ashwal and Burke1989; Moucha and Forte, Reference Moucha and Forte2011), which have further implications for erosion and sedimentation, in particular along continental margins (Lavier et al., Reference Lavier, Steckler and Brigaud2001).
3.2.1 Swells
Mantle swells likely form as a result of increased mantle heating and buoyancy of mantle plumes, and/or convergent mantle circulation systems, and can be identified from large scale deep seismic and gravity surveys (Gurnis et al., Reference Gurnis, Mitrovica, Ritsema and van Heijst2000; James et al., Reference James, Fouch, VanDecar and van der Lee2001), including identifying large low shear velocity provinces (LLSVPs) (Burke et al., Reference Burke and Gunnell2008; Austermann et al., Reference Austermann, Kaye, Mitrovica and Huybers2014). The net result of these processes is domal lithospheric uplift, which can likely be persistent over 107–108 years. The ‘African Superswell’ located beneath eastern and southern Africa (Fig. 3.1) has elevated the land surface to average values of ~1000 m above sea level (asl), compared to typical cratonic values of 400–500 m asl (Lithgow-Bertelloni and Silver, Reference Lithgow-Bertelloni and Silver1998). What is unclear is the detailed spatial pattern of mantle swells beneath Africa, their relative sizes, and their patterns of temporal evolution. This is because they may change in size and position due to changes in mantle processes (Moucha and Forte, Reference Moucha and Forte2011), and unroofing by increased land surface erosion over topographic highs can lead to isostatic rebound effects that can encourage differential uplift. The proposed stability of LLSVPs (Mulyukova et al., Reference Mulyukova, Steinberger, Dabrowski and Sobolev2015) suggests that the spatial distribution of swells beneath Africa is an outcome of plate movement over LLSVP and associated mantle plumes (Steinberger and Torsvik, Reference Steinberger and Torsvik2012). This in turn has implications for the formation of Large Igneous Provinces (LIPs), including the Karoo volcanics, that are associated with the activity of these plumes (Torsvik and Cocks, Reference Torsvik and Cocks2013). Formation of the Karoo LIP during the Toarcian (early Jurassic) is thought to have led to significant impacts on global climate, biodiversity, marine anoxia and geochemical cycling (Sell et al., Reference Sell, Ovtcharova, Guex, Bartolini, Jourdan, Spangenberg, Vicente and Schaltegger2014), and more regional impacts on weathering patterns and long-term sediment yield across southern Africa (Burke and Gunnell, Reference Burke and Gunnell2008). Moreover, the presence of acid Karoo volcanic rocks has exerted an impact on soil, vegetation and hydrological properties throughout the subsequent Cenozoic.
Fig. 3.1. Map of Africa’s major mantle swells (dark shading), rifts, and sedimentary basins (light shading). Basin abbreviations are: Taoudeni (Ta), Niger (Ni), Iullemeden (Iu), Chad (Ch), Sudd (Su), Congo (Co), Kalahari (Ka)
Several studies have examined the impacts of differential land surface uplift along the axes of adjacent swells, which are also related to phases of rifting and volcanism. These impacts are well expressed in the migration of major drainage divides, which can be evaluated through changes in sediment provenance (Moore et al., Reference Moore, Blenkinsop and Cotterill2009). The relationship between uplift of the land surface by mantle swells and development of the erosional African Surfaces was discussed by Burke and Gunnell (Reference Burke and Gunnell2008). They argued that a low-lying and weathered early Cenozoic African land surface was uplifted by the Afar (Ethiopian) mantle plume at ~31 Ma. Subsequently, differential weathering and erosion have partly preserved and partly removed this surface, leaving bauxite and laterite deposits.
Tinker et al. (Reference Tinker, de Wit and Brown2008) present thermochronometric data from the southern Cape that suggest high denudation rates during the mid to late Cretaceous (80–100 Ma) compared to low rates during the Cenozoic. The Cretaceous denudation episode may have been triggered by epeirogenic uplift associated with emplacement of LIPs. This evidence, along with similar thermochronometric evidence elsewhere across southern Africa (e.g. Brown et al., Reference Brown, Summerfield and Gleadow2002) contrasts with the viewpoint of Burke (Reference Burke1996) and Burke and Gunnell (Reference Burke and Gunnell2008) that much of the topographic evolution of southern Africa postdates the timing of the Afar mantle swell at ~31 Ma. Although land surface denudation would have continued throughout geological history, the available data are spatially sparse, capture different time slices, and likely hide spatial and temporal complexity in terms of landscape responses to mantle plumes and swells.
3.2.2 Rifts
Rifting arises as a result of plate movement, shearing, continental breakup and causes the development of ocean basins (Paton, Reference Paton2006). Stages of rifting have been summarised from across Africa by Bumby and Guiraud (Reference Bumby and Guiraud2005) with significant African rifts shown in Figure 3.2. Marginal upwarping is a characteristic property of rifting; in southern Africa this is estimated to have contributed to ~600 m of uplift compared to the continental interior (Gilchrist and Summerfield, Reference Gilchrist and Summerfield1990). This differential uplift is an isostatic response to differential denudation rates, and fits with wider discussion of forcing–response relationships to land surface uplift. For example, Matmon et al. (Reference Matmon, Bierman and Enzel2002) argued that uplift of rifted continental margins takes place only during initial stages of ocean basin development. In this model, the Great Escarpment of southern Africa does not experience significant and uniform parallel slope retreat over time, but the escarpment edge experiences an increase in sinuosity (indentedness of the scarp line position in plan view) (e.g. van der Beek et al., Reference van der Beek, Summerfield, Braun, Brown and Fleming2002). The denudation response of creation of relief by rifting can be seen in the adjacent sediment record: thick sequences (<600 m thickness) of synrift sediments covering the last 14.5 Ma are preserved in Uganda, recording both rifting and climatic phases along the southern Albert Rift system (Roller et al., Reference Roller, Hornung, Hinderer and Ssemmanda2010). Foster and Gleadow (Reference Foster and Gleadow1992) argue a direct causal relationship between climate and denudation rates of rift-margin mountains. However, this does not take into account changes in accommodation space within rift valleys. Periods of low accommodation space and/or arid climatic conditions are shown by the development of chemical weathering products (iron, carbonate and gypsum deposits) within palaeosols in these rift valley sequences.
Fig. 3.2. Schematic relationship between rifts, drainage patterns, and deltaic deposition in adjacent oceans
3.3 Dating the African land surface
Several different approaches have been used for dating the age and dynamic behaviour, such as denudation history, of the African land surface (see Woodborne, this volume). Most commonly, cosmogenic exposure or burial age dating, or thermochronometric dating, has been used. Cosmogenic radionuclide dating can inform on the timing of exposure of rock surfaces following erosion, or the timing of burial of rock surfaces by eroded sediments. These related applications of the cosmogenic method can be used to calculate point-specific and averaged rates of land surface denudation and aggradation over (generally) long timescales, commonly based upon one single radiometric age with relatively wide errors. The validity of such calculated rates has not been discussed in the literature, but regional-scale models of land surface denudation history have been based on such very limited data (e.g. Flemming et al., Reference Fleming, Summerfield, Stone, Fifield and Cresswell1999; Kounov et al., Reference Kounov, Niedermann, de Wit, Viola, Andreoli and Erzinger2007). Generally, these cosmogenic dating approaches suggest that denudation rates are both climatically and geologically dependent, with areas of high weathering rates promoting high sediment production and slope sediment transport (Kounov et al., Reference Kounov, Niedermann, de Wit, Viola, Andreoli and Erzinger2007; Chadwick et al., Reference Chadwick, Roering, Heimsath, Levick, Asner and Khomo2013; Decker et al., Reference Decker, Niedermann and de Wit2013). The thermochronometric approach, using apatite fission track dating, is based on calculated palaeotemperature and thus depth of emplacement of host igneous rocks, and reconstructed thermal history. In turn, this can be used to calculate the depth of land surface erosion. However, this method is complicated by variations in epeirogenic uplift and mantle buoyancy (Tinker et al., Reference Tinker, de Wit and Brown2008; Flowers and Schoene, Reference Flowers and Schoene2010).
The aggradational products of land surface weathering and erosion, including slope, fluvial and aeolian sediments, generally accumulate more episodically and under specific environmental conditions, when compared to the weathering processes that create them. This means that there is a mismatch between the timing and sediment yield of weathering, and the timing and sedimentation rate of aggradation. Some studies have suggested that this mismatch may be related to transient sediment storage within river basins (e.g. Coulthard and Van de Wiel, Reference Coulthard and Van de Wiel2013). It is also likely that sediment volumes within subsiding sedimentary basins or in the offshore zone are poorly estimated, which makes it difficult to backstrip these sediment volumes in order to calculate past basin accumulation rates or fluvial sediment yield.
3.4 Landscape responses to Cenozoic climate changes
Climate variability during the Cenozoic of southern Africa has been described in several studies (e.g. Partridge et al., Reference Partridge, Bond, Hartnady, deMenocal, Ruddiman, Vrba, Denton, Partridge and Burckle1995; Senut et al., Reference Senut, Pickford and Ségalen2009; Roters and Henrick, Reference Roters and Henrich2010). A potential problem with these studies is the extent to which site-scale data reflect climates across a wider region, and whether significant climatic gradients existed in the past as they do today. Irrespective of Cenozoic climatic conditions, it is likely that the land surface experienced spatially and temporally variable responses to this climatic forcing which cannot be adequately resolved from the existing dating controls on denudation and deposition (e.g. Champagnac et al., Reference Champagnac, Valla and Herman2014).
A range of land surface responses takes place as a result of climate forcing. Within any one landscape, however, different geomorphological elements may exhibit different spatial and temporal scales of response, and be affected by different time lags, feedbacks and geomorphological thresholds (Phillips, Reference Phillips2009; von Elderfeldt, Reference Erlanger, Granger and Gibbon2012; Knight and Harrison, Reference Knight and Harrison2013). This means that climate forcing across any one landscape can lead to spatial and temporal variations in geomorphic responses. The most useful ways to measure these geomorphic responses are through (1) variations in land surface radiometric ages and averaged denudation rates, and/or (2) variations in sediment yield on slopes and within river basins. Both these approaches have been used in order to evaluate the geomorphic impacts of Cenozoic climate changes in southern Africa.
3.4.1 Erosional responses: the African Surfaces
The different African Surfaces, formed as a result of peneplanation following tectonic uplift, are presented and discussed by Knight and Grab (this volume). Land surfaces that have been previously correlated to the same African Surface level show a range of ages and altitudes (Partridge and Maud, Reference Partridge and Maud1987), thus their grouping is somewhat arbitrary. For example, uplift during the Pliocene was associated with the formation of the Post-African I Surface, but the uplift ranged from 100 m in the west to 900 m in the east of southern Africa (Partridge and Maud, Reference Partridge and Maud1987, Reference Partridge, Maud, Partridge and Maud2000; Moon and Dardis, Reference Moon, Dardis, Moon and Dardis1988). In turn, this differential uplift resulted in very different land surface responses with respect to river incision and sediment yield, giving rise to major deposition in the Kalahari basin at this time. Much emphasis has been given to river system responses to uplift, with studies focusing on drainage basin evolution and migration (e.g. Moore and Larkin, Reference Moore and Larkin2001; Moore and Blenkinsop, Reference Moore and Blenkinsop2006) and changes in river long profiles (Partridge and Maud, Reference Partridge and Maud1987, Reference Partridge, Maud, Partridge and Maud2000). These viewpoints focus mainly on erosional responses, yet river deposition/aggradational responses and slope instability responses are equally important, but have not really been considered within the context of climate- or tectonics-driven sediment supply (Wintle et al., Reference Wintle, Botha, Li and Vogel1995; Botha and Partridge, Reference Botha, Partridge, Partridge and Maud2000).
3.4.2 Residual deposits
The relationship of residual (weathered) deposits to long-term landscape evolution and climate change in Africa has been described by several authors (e.g. Marker and Holmes, Reference Marker and Holmes1999; Pickford et al., Reference Pickford, Eisenmann and Senut1999; Marker et al., Reference Marker, McFarlane and Wormald2002). The relationship between the development (in thickness and spatial extent) of residual deposits and their impacts on land surface denudation is poorly known and has not been adequately resolved in any model (see Faniran and Jeje, Reference Faniran and Jeje1983). For example, chemical weathering under the typically hot, humid conditions found in many parts of central and southern Africa results in kaolinisation, leaching of silica and bases and enrichment of iron and aluminium. Such in situ weathering followed by chemical translocation may lead to negative feedbacks such as the development of iron hardpans that progressively inhibit translocation and interflow. Likewise, development of silcrete and calcrete at or just beneath the land surface can progressively decrease net denudation rates over time. For example, Gunnell (Reference Gunnell2003) suggested that bauxitisation in West Africa took place around 45–50 Ma and one of several phases of lateritisation around 24–25 Ma. These estimates were based on calculated denudation rates of ~2 m/Ma on lateritic plateaus and 7–13 m/Ma on non-lateritic surfaces. Here, the presence of a laterite cap is suggested to decrease denudation by a factor of 4–6, but it is likely that such differences in denudation rates are time-dependent. Thus, such denudation rates are not appropriate for the entire time period over which they are calculated. Several studies argue that weathering-limited denudation is the correct theoretical approach to understand landscape responses to Cenozoic climate changes (e.g. Decker et al., Reference Decker, Niedermann and de Wit2011, Reference Decker, Niedermann and de Wit2013; Chadwick et al., Reference Chadwick, Roering, Heimsath, Levick, Asner and Khomo2013), but none has considered the temporal evolution of weathering and denudation rates. Enrichment of surface minerals by lateritisation has been argued to have taken place preferentially on the African Surfaces that were subject to long periods of weathering (Beukes et al., Reference Beukes, van Niekerk and Gutzmer2004; Prendergast, Reference Prendergast2013).
3.4.3 Depositional responses: fans
Footslope deposits of alluvial and colluvial fans are key elements in the landscape history of southern Africa, because they can often be dated, and may occur on different scales from microscale footslope wedges to megafans (Tooth and McCarthy, Reference Tooth and McCarthy2007). The smaller systems (<1 ha) form wetlands in valley confluence or piedmont settings with high river water and sediment supply, and are late Pleistocene or Holocene in age (e.g. Grenfell et al., Reference Grenfell, Ellery and Grenfell2009). The larger systems were in nearly all cases formed during the mid-Cenozoic when pluvial conditions led to the formation of large lakes in southern Africa, including Palaeolake Etosha and Palaeolake Makgadikgadi (Namibia/Botswana). These lakes offered shallow water conditions with migrating channels, islands and wetlands, and formed as endorheic basins fed by rivers which had their catchments in nearby highland regions. The Etosha and Cubango megafans formed under these conditions (Lindenmaier et al., Reference Lindenmaier, Miller, Fenner, Christelis, Dill, Himmelsbach, Kaufhold, Lohe, Quinger, Schildknecht, Symons, Walzer and van Wyk2014), and now have a high preservation potential underneath a younger and relatively inactive surface, where aridification during the Neogene and Quaternary resulted in abandonment and preservation of the megafan form (Miller et al., Reference Miller, Pickford and Senut2010) and partial drying out of fans such as the Okavango Delta.
The contemporary sedimentary system of the Okavango Delta (Botswana) is the uppermost part of a much larger relict, buried megafan (McCarthy, Reference McCarthy2013). Seismic profiling data show that this is around 150 m thick and was formed in the very large (>90,000 km2) Palaeolake Makgadikgadi (Podgorski et al., Reference Podgorski, Green, Kgotlhang, Kinzelbach, Kalscheuer, Auken and Ngwisanyi2013, Reference Podgorski, Green, Kalscheuer, Kinzelbach, Horstmeyer, Maurer, Rabenstein, Doetsch, Auken, Ngwisanyi, Tshoso, Jaba, Ntibinyane and Laletsang2015). Sediments progressively infilled this sag basin during the Cenozoic (Haddon and McCarthy, Reference Haddon and McCarthy2005). Internally, Okavango sediments comprise mainly sand, silt and mud interbeds, typical of a shifting sandy braidplain delta with seasonal variations in water fluxes.
Larger fan systems that are much older and more subdued in surface relief are also found, often in piedmont settings where sediments accumulate due to a decrease in flow energy. In the Witteberg of the Cape Fold Belt, elements of the land surface of a relict alluvial fan were dated by cosmogenic methods to ~2 Ma (Kounov et al., Reference Kounov, Nierdermann, de Wit, Codilean, Viola, Andreoli and Christl2014). Many other relict alluvial fans are observed in the Richtersveld and Cederberg (Western and Northern Cape), although these have not been studied in detail.
3.4.4 Marine deposition
The activity of sediment transport systems inland often contributes to sediment supply to marine basins or shelf break canyons. For example, climatic changes from the Cretaceous to Cenozoic resulted in phases of marine sedimentation driven by erosional cycles inland (Fig. 3.2). This hints at, first, the longevity of many river systems (Moore et al., Reference Moore, Blenkinsop and Cotterill2012) and, second, the macroscale connections between different parts of integrated sediment systems that operate on a continental scale (Macgregor, Reference Macgregor2010). Modelling shows that sediment fluxes are positively enhanced during periods of tectonic uplift and tilting inland (Braun et al., Reference Braun, Guillocheau, Robin, Baby and Jelsma2014), but changing land surface topography can result in migrating watersheds and alterations in the sediment balance within individual catchments (Macgregor, Reference Macgregor2010). For example, the Congo basin appears to have received sediment from a number of different sources throughout its history (Giresse, Reference Giresse2005). Offshore and onshore patterns of basinal sedimentation may record slightly different things related to both tectonics and climate, as a result of shifting drainage divides, changing accommodation and endorheic sag basins on land, and rifting and basin subsidence in the ocean (Lavier et al., Reference Lavier, Steckler and Brigaud2001; Guillocheau et al., Reference Guillocheau, Rouby, Robin, Helm, Rolland, Le Carlier de Veslud and Braun2012). The sediment record of marine basins may therefore have a more complex and longer history than previously considered. Over shorter Quaternary timescales, sediment dynamics of the Orange River shelf responded to glacial–interglacial changes in sea level and turbidite activity (Compton and Wiltshire, Reference Compton and Wiltshire2009).
3.5 Discussion
Most of the existing landscape evolution models for southern Africa focus on tectonic forcing followed by steady weathering/geomorphic responses (Cotterill and de Wit, Reference Cotterill and de Wit2011; Kounov et al., Reference Kounov, Viola, Dunkl and Frimmel2013; Champagnac et al., Reference Champagnac, Valla and Herman2014). These models are theoretically similar to the model presented by du Toit (Reference du Toit1933) some 80 years earlier.
The events described earlier, however, represent regional- to continental-scale drivers of landscape change over specific timescales, and should be qualified in two ways. First, tectonic events prior to the Cenozoic, including linked patterns of mantle plume migration, land surface uplift and erosional responses, are known to have taken place during the late Cretaceous (Braun et al., Reference Braun, Guillocheau, Robin, Baby and Jelsma2014). These are very likely to have influenced both lithospheric relaxation and topographic evolution over 106-year timescales, and thus well into the Cenozoic. Swell-driven uplift, subsidence and migration cannot be easily reconciled with simple and deterministic patterns of tectonic uplift and lithospheric relaxation (e.g. Erlanger et al., Reference Erlanger, Granger and Gibbon2012; Kounov et al., Reference Kounov, Viola, Dunkl and Frimmel2013; Bierman et al., Reference Bierman, Coppersmith, Hanson, Neveling, Portenga and Rood2014). Second, several studies have directly examined the relationships between tectonics, climate, and land surface responses (Fig. 3.3), mainly through examining topographic (relief) and elevational (land surface height) changes, using different dating and analytical methods (e.g. Decker et al., Reference Decker, Niedermann and de Wit2013; Kounov et al., Reference Kounov, Viola, Dunkl and Frimmel2013; Champagnac et al., Reference Champagnac, Valla and Herman2014), geologic patterns of drainage evolution (Moore, Reference Moore1999; Moore et al., Reference Moore, Blenkinsop and Cotterill2012), or climate modelling approaches (Sepulchre et al., Reference Sepulchre, Ramstein, Fluteau, Schuster, Tiercelin and Brunet2006; Jung et al., Reference Jung, Prange and Schulz2014). These studies are somewhat supported by dating evidence which suggests that the southern African land surface is not of uniform age (Bierman et al., Reference Bierman, Coppersmith, Hanson, Neveling, Portenga and Rood2014). However, more widely it illustrates the potential of the southern African landscape as a palimpsest comprised of different landscape elements of different ages (Cotterill and de Wit, Reference Cotterill and de Wit2011), and exhibiting different sensitivities to climatic and associated geomorphic forcing (Knight and Grab, this volume).
Fig. 3.3. Illustration of the modelled relationship between mantle uplift and land surface tilting, and associated response through predicted sediment flux (modified from Braun et al., Reference Braun, Guillocheau, Robin, Baby and Jelsma2014). Note that tilted land surfaces provide higher initial sediment yields, but that these values are exceeded by sediment yields from non-tilted land surfaces after around 30 Ma.
A number of future research strands can be developed to better evaluate the role of tectonic and climatic events on Cenozoic landscape evolution in southern Africa. These may include: (1) using geophysical methods to evaluate weathering depth, (2) using geochemical and radiometric methods to evaluate sediment source areas and dynamics within river and slope systems, (3) intensive cosmogenic dating of land surfaces in order to evaluate the range of land surface ages within any one area, and (4) testing the concept of African Surfaces through digital hypsometry.
3.6 Summary
The southern African landscape shows complex patterns of evolution during the Cenozoic, comprising tectonic, climatic and geomorphic interactions and feedbacks. Landscape evolution models and radiometric dating of the land surface and its weathering products provide some control on the timings of different events, but no coherent spatial or temporal patterns have yet emerged. Integrated tectonic, geophysical, geochemical, geomorphological and dating studies may illuminate on some of these issues, but this requires considerable multidisciplinary research effort.
