8.1 Introduction

Many of the southern African proxy records discussed in this book are valuable in their ability to provide a chronological sequence of climatic and environmental events over periods spanning up to several thousands of years. To this end, Quaternary-age glacial and periglacial landforms are restricted, as favourable climatic conditions have only been present during narrow time-windows. Although there is considerable debate over past climatic conditions in upland areas of southern Africa where glacial and periglacial landforms would have developed (e.g. Partridge et al., Reference Partridge, Scott and Hamilton1999; Holmgren et al., Reference Holmgren, Lee-Thorp, Cooper, Lundblad, Partridge, Scott, Sithaldeen, Talma and Tyson2003; Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab, Rea, Carr and Farrow2012), these landforms are difficult to date precisely and may have developed over multiple cool climate phases, such that their preserved morphologies cannot be taken uncritically as a measure of climate severity. Notwithstanding such difficulties, correct geomorphological interpretation and dating can allow for inferences of palaeoclimate and palaeoenvironmental conditions.

A wide range of relict Quaternary glacial and periglacial landforms exists in the Western and Eastern Cape mountains, and the Drakensberg sector of the Great Escarpment which is located in the border area between Lesotho and KwaZulu-Natal Province in South Africa (Fig. 8.1). However, there is disagreement as to the origin, climatic context and age of many of these landforms. This chapter reflects on (1) the historical development of understanding Quaternary-age glaciation in southern Africa, and hypothetical concerns of such a proposed glaciation; (2) the distribution and morphology of glacial and periglacial landforms; and (3) their palaeoclimatic and palaeoenvironmental implications. This chapter highlights the changing role of the cryosphere (sensu lato) in the development of southern Africa’s high mountain landscapes, and that cold-climate processes are becoming less significant as a consequence of regional climate changes in the Anthropocene.

Fig. 8.1. (A) Location of key mountainous areas in southern Africa, and (B) close up of the study area (boxed in part A) showing the locations of places named in the text.

8.2 Setting the scene

The earliest glacial evidence in southern Africa dates to the Permo-Carboniferous when the African subcontinent was located farther south as part of Gondwanaland. This glacial episode is manifested in the Dwyka Group of the Karoo Sequence, consisting mainly of glaciomarine deglacial sequences which in places include boulder or conglomeratic tillites (Visser and Hall, Reference Visser and Hall1985; von Brunn, Reference von Brunn1996; Isbell et al., Reference Isbell, Cole, Catuneanu, Fielding, Frank and Isbell2008). Surface-exposed glacially striated pavements with crescentic, circular and oval shaped markings and fractures have helped identify former ice flow patterns and calculate the original thickness of Karoo strata (Master, Reference Master2012). The Dwyka Group sediments extend over ~600,000 km² and attain a maximum thickness of ~800 m along the southern margins of the basin, but thin northwards to between 0 and 600 m thick (Isbell et al., Reference Isbell, Cole, Catuneanu, Fielding, Frank and Isbell2008). This early glacial evidence, from across large areas of southern Africa, is important to understand late Paleozoic glaciations, climate dynamics, and continental drift (e.g. du Toit, Reference du Toit1937). Significant evidence exists, however, for the impacts of cold-climate (glacial and periglacial) processes in southern Africa during the Quaternary. In this chapter, evidence for late Pleistocene-age glaciation and periglaciation, within the Quaternary period, is considered from two regions in southern Africa (the Drakensberg and southwestern Cape).

In the Drakensberg sector of the Great Escarpment (between 28˚30′ and 31°S), fluvial and aeolian deposition continued through the Triassic and Jurassic, laying down a succession of variably textured sediments of the Karoo Supergroup (Carboniferous to Jurassic), and culminating in the fine-grained Clarens Formation which reflects increased aridity and sand dune deposition (Holzförster, Reference Holzförster2007). These geological processes were important to develop sufficient relief to induce cold climate conditions on mountain tops. The sediments were then overlain by the relatively short-lived Drakensberg Group volcanic event at ~182 Ma (Jourdan et al., Reference Jourdan, Féraud, Bertrand, Watkeys and Renne2007). Strike-slip faulting and rapid denudation of the continental margins during continental rifting of the early Cretaceous (~136 Ma) is manifested today as the Great Escarpment, located up to several hundred kilometres inland of the contemporary continental margin. In the Drakensberg sector, mountain peaks rise to over 3400 m a.s.l. – the highest in southern Africa – which made them the most climatically susceptible locations for Quaternary glaciation.

In the southwestern Cape, the Cape Supergroup (Ordovician to Devonian) comprises a 6–10 km thick sandstone sequence, deposited along a subsiding shelf. Subduction and accretion of the palaeo-Pacific plate beneath Gondwana subsequently caused regional folding of the sandstone strata around 330 Ma, forming the Cape Fold Belt (Booth, Reference Booth2011). These mountains today extend east–west across the southern Eastern Cape and Western Cape provinces (Swarteberge, Outeniqua and Langeberg ranges; between ~25 and 19°E), and south–north in the Western Cape (Cederberg range, between ~34 and 31°S). The highest peaks reach up to 2325 m a.s.l. Here, the Great Escarpment is located 140–220 km farther inland, and has a lower maximum altitude (~1600 m a.s.l.) than the Cape Fold Mountains (Fig. 8.1). The highest peaks here are mainly underlain by resistant sandstones and quartzites.

8.3 Proponents for southern African glaciation during the late Pleistocene: a brief history

In the 1940s and 1950s, several workers considered the possibility of Quaternary glaciations in southern Africa (e.g. Kojot, Reference Kojot1948; du Toit, Reference du Toit1954), but it was generally thought that the climate would have been too dry to sustain any glaciers. Flint (Reference Flint1959) proposed that late Pleistocene glaciers did exist in southern Africa, but were confined to the highest mountains. Sparrow (Reference Sparrow1964, Reference Sparrow1967) identified slope elements consistent with glaciation, such as U-shaped valleys, but these were localised and not compatible with any widespread glaciation. Subsequently, he reinterpreted such upland hollows and valleys as ‘non-glacial cirques’ formed by enhanced rock surface weathering under a range of climatic conditions (Sparrow, Reference Sparrow1974). Marker and Whittington (Reference Marker and Whittington1971) examined the morphometry of 577 slope hollows located at high elevation positions (>2900 m a.s.l.) in the Drakensberg sector of the Great Escarpment, in Lesotho. They found hollows to have a consistent morphology, spatial variation and a preferred north-facing aspect (see Dyer and Marker, Reference Dyer and Marker1979). The work was later expanded to cover 12 different morphometric parameters tested upon 628 hollows covering 156,000 km² of the Lesotho Highlands (Marker, Reference Marker1991). Cold Pleistocene climate conditions and associated glaciations were thought to have been the main driver for hollow development (Marker, Reference Marker1991).

In the Cape Fold Belt, evidence for proposed plateau, cirque and valley glaciation during the Pleistocene was based on morphometric air photo analysis (cirques and moraines) and rock surface observations (glacial polishing and striae) (Borchert and Sänger, Reference Borchert and Sänger1981; Sänger, Reference Sänger, Matheis and Schandelmeier1987). Sänger (Reference Sänger1988) proposed a snowline depression to ~1700 m a.s.l. during the Last Glacial Maximum (LGM) over the Hex River mountains, which are located in the direct path of moisture-laden mid-latitude cyclones approaching from the southwest. Several subsequent studies, based on a range of proxy indicators, have suggested a northerly displacement of pressure belts during the LGM (e.g. Gellert, Reference Gellert1991; Gasse et al., Reference Gasse, Chalié, Vincens, Williams and Williamson2008), which would indeed have increased snowfall across southeastern African mountains.

Given the problems of misidentification, and the attribution of glaciation to erosional features alone, especially when based on small scale phenomena such as striae, more recent studies have focussed on identifying and interpreting possible depositional evidence for glaciation. Several studies have examined apparent terminal moraine ridges that may correspond to cirque glaciers in the Drakensberg sectors of the Eastern Cape (Lewis and Illgner, Reference Lewis and Illgner2001) and KwaZulu-Natal/Lesotho Highlands (Mills and Grab, Reference Mills and Grab2005; Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb). The elevations of these moraine ridges have been used to calculate glacier equilibrium line altitudes (ELAs) and thus inform on LGM climate (Mills et al., Reference Mills, Grab, Rea, Carr and Farrow2012).

8.4 The opponent view to southern African glaciation during the late Pleistocene

The geomorphic evidence used to argue for past Quaternary glaciations in southern Africa has been challenged and debated at length over the years (see Boelhouwers and Meiklejohn, Reference Boelhouwers and Meiklejohn2002; Hall, Reference Hall, Ehlers and Gibbard2004; Osmaston and Harrison, Reference Osmaston and Harrison2005). Opponent views have challenged the scientific rigour and theoretical basis of Quaternary glaciations, although largely without alternative hypotheses based on field evidence (Knight, Reference Knight2012). For example, the Western Cape ‘moraines’ have apparent bedrock remnants within them and thus are likely structural phenomena, while striated pebbles are argued to be associated with Silurian-age Pakhuis Tillites (Hall, Reference Hall, Ehlers and Gibbard2004). There is also uncertainty on the climatic conditions of the LGM, with most discussion focusing on the precipitation (snow) required to sustain glaciers. There is a lack of independent evidence from other climate proxies to precisely evaluate precipitation conditions, but changes in synoptic climatological patterns (and thus moisture sources) are also uncertain (cf. Chase and Meadows, Reference Chase and Meadows2007). Arguments for and against the presence of Quaternary glaciers invoke high and low precipitation values, respectively, but both viewpoints are based on uncritical circular arguments in which, for example, the interpreted presence of a former glacier is assumed to also imply cold temperatures and wet conditions, even if there is no independent evidence for this (see Boelhouwers and Meiklejohn, Reference Boelhouwers and Meiklejohn2002; Hall and Meiklejohn, Reference Hall, Meiklejohn, Ehlers and Gibbard2011). Therefore, the climatology of the Pleistocene LGM must be considered independently of any inferences of climate at this time, based on geomorphological evidence. In the Drakensberg, depositional features interpreted as terminal moraines have been used to suggest the presence of late Pleistocene glaciers. This evidence and its implications are now discussed.

8.5 Glacial depositional landforms in the Drakensberg

Some depositional landforms (mainly debris ridges) have been identified in several Drakensberg locations, and interpreted as terminal moraines. Such ridges, however, have as yet only been reported from four sites: Mount Enterprise (~31˚11′00″S, 28˚00′00″E), Leqooa Valley (~29˚44′35″S, 29˚06′57″E), Sekhokong Range (29˚36′51″S, 29˚13′51″E) and Tsatsa-La-Mangaung Range (29˚33′35″S, 29˚17′39″E) (Lewis and Illgner, Reference Lewis and Illgner2001; Mills and Grab, Reference Mills and Grab2005; Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb) (Fig. 8.1). The Mount Enterprise site is located in the southernmost end of the Drakensberg range, Eastern Cape Province. Here, the maximum summit altitude is 2565 m a.s.l., but escarpment cliffs generally terminate between 2200–2300 m a.s.l. The moraine ridges located at more northerly sites are all found in eastern Lesotho, on summit sites close to the Great Escarpment edge. The mountains west (inland) of the Great Escarpment also have summits exceeding 3300 m a.s.l., but ridges interpreted as glacigenic are conspicuously absent north of the Tsatsa-La-Mangaung Range.

All of the identified moraine ridges in eastern Lesotho, from different locations, occur within a narrow altitudinal range of 3005–3112 m a.s.l., which suggests a strong climatic control on their formation and thus does not favour alternative interpretations such as slope failure or colluviation (Fig. 8.2). However, the more southerly Mount Enterprise ridges are located between 2040–2059 m a.s.l. Notably, this region today receives considerably more frequent and heavier snowfalls than the higher altitude sites to the north. Maximum summit elevations (and calculated ELAs) above sea level are 3257 m (3095 m) at Sekhokong; 3431 m (3298 m) at Leqooa (Mills et al., Reference Mills, Grab, Rea, Carr and Farrow2012); and 2302 m (2109 m) at Mount Enterprise (Lewis and Illgner, Reference Lewis and Illgner2001) (Fig. 8.3). The large ELA differences between Lesotho and adjoining (within ~2˚ latitude) areas of the Eastern Cape beg questioning. This difference may imply possible misidentification of some landforms, or substantial palaeoclimatic gradients over southeastern Africa, which is not the case today, even though synoptic circulation patterns during the LGM were likely different to present (e.g. Cockcroft et al., Reference Cockcroft, Wilkinson and Tyson1987; Mills et al., Reference Mills, Grab, Rea, Carr and Farrow2012). At the sites in Lesotho, moraines are between 100–300 m in length, 15–70 m in width and 4–16 m in height. Thus, the glaciers would have all been small (0.05–0.31 km²), located well below adjacent summits, with limited accumulation areas, and a mass balance strongly reliant on snowblow (Mills et al., Reference Mills, Grab, Rea, Carr and Farrow2012). Lack of sedimentological exposures through the ridges has limited analysis, but where trenched, the ridges generally consist of poorly sorted diamictons composed of clasts in a finer matrix. The covariance plot of C₄₀ and RA indices fits within the morainic sediment envelopes according to Benn and Ballantyne (Reference Benn and Ballantyne1994) (see Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb). These macro-sedimentological characteristics likely support both passive and active glacigenic transport modes (Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb). Micromorphological analyses from the Lesotho sites display turbate structures, ‘halo’ structures and pressure shadows, which imply particle rotation during transport. The presence of in situ fractured quartz grains indicates high normal stress conditions, and subglacial stresses in particular (Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb). Farther south, the Mount Enterprise ridges consist of clast- and matrix-supported clasts, some of which are striated (Lewis and Illgner, Reference Lewis and Illgner2001). Although the sediment coincides with particle-size distribution envelopes associated with glacial lodgement and melt-out tills (see Sladen and Wrigley, Reference Sladen, Wrigley and Eyles1983), no macro-sedimentological analyses have as yet been undertaken which may either support or refute a glacial origin.

Fig. 8.2. Dissected glacigenic ridge forms, Leqooa valley, Lesotho (photo reproduced courtesy of Stephanie Mills). The ridges are buttressed (centre of photo) against bedrock slopes behind.

Fig. 8.3. Graph showing the relationship between ELA value and location of glacial sites in southern Africa. The Lesotho Highlands sites (two from the Sekhokong Range, two from Leqooa Valley, one from Tsata-La-Mangaung Range) show internal consistency.

8.6 Palaeoclimate implications of late Pleistocene glaciation in southeastern Africa

Radiocarbon-dated soil organic matter from within five moraines in eastern Lesotho provides some indication of minimum moraine ages. Calibrated ages range between 13,820–20,530 cal yr BP, and confirm ridge deposition during the last glacial period (Mills et al., Reference Mills, Grab and Carr2009a). Although no reliable dates are available for the Mount Enterprise deposits, these are assumed to ‘be of Last Glacial Maximum or younger age’ (Lewis and Illgner, Reference Lewis and Illgner2001, p370). A caveat, however, is that a temperature depression of 17˚C below present values would be required in order to generate a LGM cirque glacier at Mount Enterprise (Lewis and Illgner, Reference Lewis and Illgner2001), which does not correspond with other regional palaeotemperature proxies for this period.

To generate small glaciers, climate models are commonly tuned to different values (relative to present) of palaeotemperature and palaeoprecipitation. By these means, the most likely climatic conditions required to generate and sustain a small glacier can be established. Based on regional climate reconstructions from a variety of proxy records (Tyson and Partridge, Reference Tyson, Partridge, Partridge and Maud2000), Mills and Grab (Reference Mills and Grab2005) plotted likely LGM summer temperature (controlling ablation) and winter precipitation (controlling accumulation) for some of the eastern Lesotho glacier sites. In contrast to the Drakensberg sector of the Eastern Cape, Mills et al. (Reference Mills, Grab, Rea, Carr and Farrow2012) calculated that with a temperature reduction of ~6˚C (as indicated by proxy data), palaeoprecipitation would have been ~1100±200 mm/pa (~0–60% higher than at present). Further, the Lesotho sites are located in topographic positions which coincide with areas of longest-lasting winter snowcover (Grab et al., Reference Grab, Mulder and Mills2009), which supports their association with palaeo-snow accumulation sites (and thus possible glacial ice).

According to Holmgren et al. (Reference Holmgren, Lee-Thorp, Cooper, Lundblad, Partridge, Scott, Sithaldeen, Talma and Tyson2003), the LGM of southern Africa experienced an average and sustained temperature depression of ~6˚C, with cooler pulses between 23–21, 19.5–17.5 and 15.0–13.5 kyr BP. Conditions during any of these periods would have favoured niche or cirque glaciers to develop in suitable topographic locations in the southern Drakensberg (Mills et al., Reference Mills, Grab and Carr2009a, Reference Mills, Grab and Carrb), as is now supported by geomorphic evidence, outlined earlier. Previous concerns that the region was too dry during the LGM to support small glaciers (Hall, Reference Hall, Ehlers and Gibbard2004) were largely based on an assumption presented by Partridge et al. (Reference Partridge, Scott and Hamilton1999) that palaeoprecipitation along the Drakensberg was reduced by ~30% during the LGM, and that there was no evidence to support wetter conditions. However, spatial patterns of palaeoprecipitation values during the LGM are now known to be spatially variable across southern Africa, based on pollen and other biological records (e.g. Gasse et al., Reference Gasse, Chalié, Vincens, Williams and Williamson2008; Scott et al., Reference Scott, Neumann, Brook, Bousman, Norström and Metwally2012). A key unknown factor, however, is the role of orography (e.g. the Great Escarpment) in modifying past precipitation patterns, including its seasonality. This means that palaeoclimate records from lowland sites might not be suitable analogues for precipitation at upland sites. Using the HadAM3h climate model and focusing on time slices at 21, 18 and 15 kyr BP, Mills et al. (Reference Mills, Grab, Rea, Carr and Farrow2012) argued that the Lesotho Highland sites would have experienced greater precipitation seasonality than at present, with more spring–winter– autumn precipitation (much of it as snow), and less summer precipitation. This provides climatic support for the development of cirque glaciers during the LGM.

8.7 Periglacial phenomena

Mountains of the Western Cape, Eastern Cape and Lesotho/Drakensberg all host contemporary and relict periglacial phenomena, and thus can be considered as the ‘core periglacial regions’ of southern Africa. Contemporary (active) periglacial features include earth hummocks (thúfur), miniature varieties of sorted patterned ground, stone- and turf-banked lobes/sheets, terraces, and soil frost phenomena associated with needle ice. Soil frost and other subsurface processes can also operate synergistically with surface processes (e.g. fluvial, aeolian) to contribute to ongoing denudation of mountain slopes across southern Africa, which is a major contemporary management issue (Grab, Reference Grab2010).

A variety of landforms interpreted as relict periglacial features has been reported from the core periglacial regions, including rock glaciers, pronival (‘protalus’) ramparts, ice wedge casts, nivation hollows, asymmetric valleys, openwork block deposits (blockslopes, blockstreams), large sorted patterned ground, thúfur, and turf-/stone-banked lobes (e.g. Lewis and Hanvey, Reference 135Lewis and Hanvey1993; Boelhouwers, Reference Boelhouwers1994; Grab, Reference Grab2000, Reference Grab2002; Sumner, Reference 136Sumner2004) (Fig. 8.4). However, in many cases the correct identification of these features is not always straightforward. Some features are still under debate in terms of their origin and significance, and the environmental and climatic interpretation of such features remains controversial (see Grab et al., Reference Grab, Mills, Carr, Holmes and Meadows2012, for a full review).

(A) Views of the blockstream near Sani Pass (as described by Boelhouwers et al., Reference Boelhouwers and Meiklejohn2002) showing its well defined margins and weathered and reoriented large clasts;

(B) stone bank lobes (arrows);

(C) sorted circle;

(D) thúfur

(photos A,D: Jasper Knight; B,C: Stefan Grab).

Fig. 8.4. Periglacial phenomena from the Lesotho Highlands.

The most convincing and prominent periglacial landforms are those associated with block deposits. Openwork block deposits in the form of blockslopes and blockstreams are relatively common and the most conspicuous large-scale cryogenic landforms around higher summits of the core periglacial regions (Fig. 8.4A). Most blockstreams occur on colder, high altitude, south-facing slopes of the high Drakensberg. Blockstream dimensions in the order of several tens of metres to ~1.1 km in length, are up to 75 m wide and typically consist of subangular blocks 0.2–1.5 m long (see Boelhouwers, Reference Boelhouwers1999; Grab, Reference Grab1999; Boelhouwers et al., Reference Boelhouwers and Meiklejohn2002). In the Matroosberg (Western Cape), blockstream clasts are platy and exhibit well-developed imbrication and downslope alignment, but in places with reorientation into a transverse fabric, dipping upslope at the blockstream frontal lobe (Boelhouwers, Reference Boelhouwers1999). The most blocky clasts within the Drakensberg blockstreams tend to have a bi- or tri-modal clast orientation pattern, with primary alignment parallel to local slope gradient (Boelhouwers et al., Reference Boelhouwers and Meiklejohn2002).

Large stone-banked lobes with tread lengths of up to 30 m and frontal banks 2–10 m wide and up to 3 m high have been reported from Drakensberg summits above 3400 m a.s.l. (Grab, Reference Grab2000) (Fig. 8.4B). In places, secondary (active) solifluction features such as stone-banked sheets have developed on the stone-banked treads (Boelhouwers, Reference Boelhouwers1994). The secondary sorting is to a depth of ~20 cm and is possibly a good indication of site-specific contemporary freeze depth. The primary lobes are sorted to depths of up to 60 cm and exhibit imbricated blocks dipping downslope.

Relict large sorted patterned ground has been reported from the Hex River Mountains, Western Cape (Boelhouwers, Reference Boelhouwers1999), and KwaZulu-Natal/Lesotho Drakensberg (Boelhouwers, Reference Boelhouwers1994; Grab, Reference Grab2002; Sumner, Reference 136Sumner2004) (Fig. 8.4C). The widespread openwork block accumulations in the Hex River Mountains host large sorted stripes up to 150 m long and 2 m wide (Boelhouwers, Reference Boelhouwers1999). In contrast, large sorted circles in the high Drakensberg contain gravel to cobble centres, surrounded by openwork borders (Grab, Reference Grab2002). Larger sorted circles are ~1–5 m in diameter and sorted to over 1 m depth. This exceeds contemporary freeze depth, which is 40–50 cm during ‘normal’ winters on the high mountain interfluves where the patterns occur (Grab, Reference Grab2004). This freezing regime also influences the formation and longevity of thúfur (earth hummocks) (Fig. 8.4D).

8.8 Discussion

In terms of reconstructing palaeoclimatic and palaeoenvironmental conditions in southern Africa during the Quaternary, the available geologic and isotopic records are highly variable in terms of length, resolution, location and environmental setting. As such, it is often difficult to compare one dataset to another or to extrapolate from one location to a wider region. This is particularly the case in mountainous regions of southern Africa where Quaternary climate records are uncommon and where climatic data from mountain-foot locations cannot be extrapolated to mountain tops. Consequently, the preserved record of glacial and periglacial phenomena often represents the only data source for Quaternary climates and environments in these mountain regions, limiting the extent to which the climatic boundary conditions of such analogues can be used.

The lack of constraining palaeoclimatic records and the absence of radiometric dating are significant limiting factors when using glacial and periglacial evidence from southern Africa. Radiocarbon dating is only possible where organic matter is present but this can provide only a minimum age of the underlying materials (see Woodborne, this volume). Cosmogenic nuclide dating of rock surfaces that have been exposed by glacier erosion does not suit the situation in southern Africa where the rocks are highly weathered and where the amount of glacier erosion is uncertain or may be minimal. The possible polygenic or multi-phase evolution of cryogenic landforms has not been fully explored, and even well-studied landforms such as blockstreams have yielded little conclusive evidence concerning their origins and palaeoenvironmental setting (e.g. Boelhouwers et al., Reference Boelhouwers and Meiklejohn2002). It is also likely that high-altitude areas of the Drakensberg and Cape Fold Belt experienced significant differences in microclimate, controlled by aspect, snowcover duration and prevailing winds. These microscale controls are very difficult to resolve with sparse field data or by downscaled climate models, and are thus a significant challenge for future research.

8.9 Summary

There remains considerable uncertainty in the status of southern African glacial and periglacial records. This is largely because the evidence for cryogenic processes and associated landform development is often ambiguous, which is partly due to their marginal climatic setting, even during the coldest parts of the Quaternary. As such, there is disagreement on the identification and interpretation of many mountain landforms. Evidence of glacial moraines in the Lesotho sector of the Drakensberg range suggests that conditions were cold and wet enough during the LGM to sustain small glaciers, with a temperature depression of ~6^oC and ELAs down to ~3100 m a.s.l. Glacier reconstruction and mass balance modelling has enabled the reconstruction of regional LGM rainfall patterns, which at this time was enhanced during autumn-winter-spring. Evidence for enhanced periglacial activity at this time is uncertain, due to the absence of secure age-determinations, but is likely to have been widespread. Better identification and age determination of diagnostic periglacial landforms are important future research strands.