The classic South African Karoo Supergroup (Du Toit 1926) is a suite of Carboniferous to Jurassic sedimentary rocks shed from the Gondwanide Mountains toward lower latitudes, infilling foreland basins, and capped by volcanics generated in conjunction with the rifting of eastern Gondwana (Cox 1970). The Karoo succession begins with Dwyka Group glacial sediments. The Dwyka is followed by the marine and coal-bearing Ecca Group, the richly fossiliferous and mainly fluviatile Beaufort Group, succeeded by the increasingly eolian Stormberg Group, and culminated by the Drakensberg Volcanics (Johnson et al. 1996; Smith et al. 1993).
U/Pb radiometric dates of 302.0 ± 3.0 Ma and 299.2 ± 3.2 Ma were determined from tuffs in the Dwyka Group (as discussed by Stollhofen et al. 2000). In the overlying Ecca Group, dates of 289.6 ± 3.8 Ma and 288.0 ± 3.0 Ma were determined from tuffs in the Prince Albert Formation (Stollhofen et al. 2000). An additional date of 270 ± 1.0 for the Ecca Group was reported from the Collingham Formation (Turner 1999). An absolute age determination of 265 ± 2.5 from the Gai-As Formation, Namibia (Hancox and Rubidge 2001), is correlated with the base of the Beaufort Group. Duncan et al. (1997) reported that Drakensberg Volcanics present highly resolved ages ranging from 184 to 179 Ma, with the majority of dates clustering at 183 ± 1 Ma.
The Carboniferous-Permian boundary falls within the Dwyka Group (Bangert et al. 1999), the Permian-Triassic boundary falls within the Beaufort Group (Hancox et al. 2002), and the Triassic-Jurassic boundary falls within the Elliot Formation of the Stormberg Group (Lucas and Hancox 2001). Increasing aridity and temperature amelioration is indicated throughout the South African succession, due in large part to the northward drift of Gondwana from a near polar 75† South latitude to a low temperate 35† South latitude (Visser 1991), but depositional styles were also influenced by tectonism during the 120 million years of Karoo deposition (Bordy et al. 2004; Smith et al. 1993).
Malawi lies in south central Africa. Its dominant geographic feature is Lake Malawi, a Rift Valley lake and the third deepest lake in the world. Karoo Supergroup rocks are present as fault block outliers in the north and the south of the country. In addition to the therapsid-bearing Chiweta Beds, bone-bearing rocks of Cretaceous age (Dixey 1928; Haughton 1928; Jacobs et al. 1993) and the Pliocene hominid-bearing Chiwondo Beds (Sandrock et al. 1999; Schrenk et al. 1993) occur in fault grabens located in close proximity to each other along the structurally complex Rift Valley in the north of the country.
The most southerly outcrops of Karoo Supergroup sediments and volcanics in Malawi are referred to as the Chikwawa Group by Habgood (1963; Figure 1) and are preserved in the Shire-Zambezi fault trough between Chikwawa and Chiromo (Habgood 1963). The capping volcanics of the Chikwawa Group extend through the Lupata Gorge area of northern Mozambique to the Lebombo Mountains of southern and western Mozambique and Swaziland, to the Drakensberg Volcanics of Lesotho and South Africa. In southern Malawi, as in northern Mozambique, Cretaceous sediments and volcanics referred to as Lupata Group (Dixie and Smith 1929) overlie Karoo volcanics.
Fragmentary bones identified as “Stormberg dinosaurs” by Sidney Haughton were reported from sediments below the Karoo volcanics in southern Malawi by Dixey (1930). The stratigraphic position of those fossils would suggest a Triassic or Early Jurassic age (Habgood 1963). Karoo volcanics are absent from northern Malawi, and none of the Karoo sediments in the north can be conclusively demonstrated to be younger than Permian. The fossils recovered from the Chiweta Beds are clearly comparable in a biostratigraphic sense to those from the Beaufort Group of the South African Karoo (Rubidge 1995), and as discussed below, indicate a Late Permian age for the Chiweta Beds.
The first bones recognized by Europeans in Karoo rocks of Malawi were fish found in 1883 by Henry Drummond in the far north of the country (Andrew and Bailey 1910; Arber 1910; Drummond 1884, 1888; Jacobs et al. 1992; Jones 1890; Newton 1910; Traquair 1910). The therapsid-bearing Chiweta Beds were visited in 1925 by Frank Dixey during reconnaissance of the area near Livingstonia and south of Mount Chombe (known as Mount Waller in older literature; Dixie 1926; Figure 2). Dixey sent bones from Chiweta to Sidney Haughton in South Africa. Within the sample, Haughton (1926) recognized gorgonopsians, of which he named Chiwetasaurus dixeyi, Dixeya quadrata, and Aelurognathus nyasaensis, and he recognized dicynodonts.
In 1930, not long after Haughton’s (1926) publication, F.W.H. Migeod spent two weeks in the Chiweta Beds with F.R. Parrington (Jacobs et al. 1992), and stated that “Dicynodon bones were the most common” fossils (Migeod 1931). Collections made by Migeod were shipped to the British Museum (Natural History).
Sigogneau (1970) and Sigogneau-Russell (1989) reviewed the Chiweta gorgonopsians originally reported by Haughton (1926). While noting that there was uncertainty as to generic identity, she transferred the species Chiwetasaurus dixeyi to Gorgonops with question. Haughton’s Aelurognathus nyasaensis, while maintained as valid, was considered close to A. trigriceps, but Sigogneau-Russell acknowledged that the species was poorly known and might belong to the Rubidgeinae rather than the Gorgonopsinae, where it was originally placed. The genus Dixeya was synonymized with Aelurognathus, thereby establishing the Chiweta species as A. quadrata.