Abstract
M.Sc. (Geology)
The south-easternmost Kaapvaal Craton is composed of scattered inliers of Archaean
basement granitoid-greenstone terrane exposed through Phanerozoic cover successions. In
addition, erosional remnants of the supracrustal Mesoarchaean Pongola Supergroup
unconformably overlay this granitoid-greenstone terrane in the same inliers. Into this crust a
variety of Precambrian intrusions occur. These are comprised of SE-, ENE- and NE-trending
dolerite dykes. Also, the Hlagothi Complex intrudes into Pongola strata in the Nkandla
region, particularly the quartzites of the basal Mantonga Formation. The whole area,
including Phanerozoic strata, has in turn been intruded by Jurassic sills and dykes related to
the Karoo Large Igneous Province. All the rocks of the Archaean inliers, with the exception of
the Jurassic sills and dykes have been subjected to greenschist facies metamorphism and
deformation, with petrographic, Ar-Ar geochronologic and palaeomagnetic studies attesting
to this. This metamorphism and deformation is associated with the Mesoproterozoic
orogeny from the nearby Namaqua-Natal Mobile Belt located to the south. This orogeny has
a decreasing influence with distance from the cratonic margin, and is highly variable from
locality to locality. However, it is generally upper greenschist facies up to a metamorphic
isograd 50 km from the craton margin. Overprints directions seen within the palaeomagnetic
data confirm directions associated with the post-Pongola granitoids across the region and
the Namaqua-Natal Mobile Belt.
The dolerite dykes consist of several trends and generations. Up to five different
generations within the three Precambrian trends have potentially been recognised. SEtrending
dykes represent the oldest dyke swarm in the area, being cross-cut by all the other
dyke trends. These dykes consist of two possible generations with similar basaltic to basaltic
andesite geochemistry. They provide evidence of a geochemically enriched or contaminated
magma having been emplaced into the craton. This is similar to SE-trending dolerite dyke swarms across the Barberton-Badplaas region to the north from literature. In northern KwaZulu-Natal the SE-trending dolerite dyke swarms have been geochronologically, geochemically and paleomagnetically linked to either ca. 2.95 or ca. 2.87 Ga magmatic events across the Kaapvaal Craton. The 2866 ± 2 Ma Hlagothi Complex is composed of a series of layered sills
intruding into Nkandla sub-basin quartzites of the Pongola Supergroup. The sills consist of
meta-peridotite, pyroxenite and gabbro. At least two distinct pulses of magmatism have
been recognised in the sills from their geochemistry. The distinct high-MgO units are
compositionally different from the older Dominion Group and Nsuze Group volcanic rocks, as
well as younger Ventersdorp volcanic rocks. This resurgence of high-MgO magmatism is
similar to komatiitic lithologies seen in the Barberton Greenstone Belt. It is indicative of a
more primitive magma source, such as one derived from a mantle plume. A mantle plume
would also account for the Hlagothi Complex and the widespread distribution of magmatic
events of possible temporal and spatial similarity across the craton. Examples include the
layered Thole Complex, gabbroic phases of the ca. 2990 to 2870 Ma Usushwana Complex,
and the 2874 ± 2 Ma SE-trending dykes of northern KwaZulu-Natal already described above
and dated herein. A generation of NE-trending dolerite dykes in northern KwaZulu-Natal can
also be palaeomagnetically linked to this event with either a primary or overprint direction.
Flood basalts seen within the upper Witwatersrand and Pongola Supergroups (i.e., Crown,
Bird, Tobolsk and Gabela lavas) may also be related. This large, voluminous extent of
magmatism allows us to provide evidence for a new Large Igneous Province on the Kaapvaal
Craton during the Mesoarchaean. This new Large Igneous Province would encompass all of
the above mentioned geological units. It is possible that it could be generated by a shortlived
transient mantle plume(s), in several distinct pulses. This plume would also explain the
development of unconformities within the Mozaan Group. This is reasoned through thermal
uplift from the plume leading to erosion of the underlying strata, culminating in the eruption
of flood basalts coeval to the Hlagothi Complex. Marine incursion and sediment deposition
would occur during thermal subsidence from the plume into the Witwatersrand-Mozaan
basin. This magmatic event also assists in resolving the apparent polar wander path for the
Kaapvaal Craton during the Meso- to Neoarchaean. Between existing poles established for
the older ca. 2.95 Ga Nsuze event, to poles established for the younger ca. 2.65 Ga
Ventersdorp event, a new magnetic component for this ca. 2.87 Ga magmatic event can be
shown. This new component has a virtual geographic pole of 23.4° N, 53.4° E and a dp and
dm of 8.2° and 11.8° for the Hlagothi Complex, with a similar magnetic direction seen in one
generation of NE-trending dolerite dykes in the region. This new ca. 2870 Ma addition to the
magmatic barcode of the Kaapvaal Craton allows for comparisons to be made to other coeval magmatic units on cratons from around the world. Specific examples include the
Millindinna Complex and the Zebra Hills dykes on the Pilbara Craton. Precise age dating and
palaeomagnetism on these magmatic units is needed to confirm a temporal and spatial link
between all the events. If substantiated, this link would assist in further validating the
existence of the Vaalbara supercraton during the Mesoarchaean.
After the Hlagothi Complex event, different pulses of magma can be seen
associated with the Neoarchaean Ventersdorp event. A generation of NE-trending dolerite
dykes in the region was dated herein at 2652 ± 11 Ma. In addition, a primary Ventersdorp
virtual geographic pole established in Lubnina et al. (2010) from ENE-trending dolerite dykes
was confirmed in this study. This ENE-trending dolerite dyke has a virtual geographic pole of
31.7° S, 13.6° E and a dp and dm of 7.0° and 7.2°. This date and virtual geographic poles
from NE- and ENE-trending dolerite dyke swarms in northern KwaZulu-Natal match up with
NE- and E-trending palaeostress fields seen in the Neoarchaean Ventersdorp and proto-
Transvaal volcanics by Olsson et al. (2010). Both generations of dolerite dykes also
demonstrate variable geochemistry. The NE-trending dolerite dyke swarm is tholeiitic, and
the ENE dolerite dyke swarm is calc-alkaline. In addition, some of the tholeiitic NE-trending
dolerite dykes have a similar magnetic component to NE-trending dolerite dykes much
further to the north in the Black Hills area according to Lubnina et al. (2010). This magnetic
component is also similar to the Mazowe dolerite dyke swarm on the Zimbabwe Craton. The
NE-trending dolerite dykes in the Black Hills area differ geochemically from those in northern
KwaZulu-Natal though, but are also of ca. 1.90 Ga age. The Mazowe dolerite dyke swarm
was linked to the dyke swarm of the Black Hills dyke swarm through palaeomagnetic studies.
The Mazowe dolerite dyke swarm however is geochemically similar to the NE-trending
dolerite dykes of northern KwaZulu-Natal, creating greater complexity in the relationship
between the three dyke swarms. It is clear from the complex array of dolerite dyke swarms
and other intrusions into these Archaean inliers of northern KwaZulu-Natal, that much more
work on the dykes within the south-easternmost Kaapvaal Craton needs to be done. This will
resolve these complex patterns and outstanding issues with regard to their palaeo-tectonic
framework.