We acquired He dates for 24 aliquots of apatites from four samples (Table DR1). These included 10 single-grain apatites from the kimberlite matrix from the Uintjiesberg, Markt and Melton Wold kimberlites, and 10 single-grain apatites from crustal clasts consisting mostly of gneissic basement from Markt and Uintjiesberg. From Hebron we acquired 2 single-grain and 2 multigrain fractions of 4 grains each for low eU apatites from an amphibolite xenolith. Individual crystals were selected based on crystal form and clarity. Grains were examined for mineral inclusions using a binocular microscope with crossed polars. Dimensions were measured from photographed grains prior to packaging in Pt packets for analysis. He measurements were made at the University of Colorado Boulder (CU) and the California Institute of Technology (Caltech). He analyses at Caltech followed the procedures outlined by House et. al. (2000). At CU, grains were heated at 6A for 5 minutes to extract the He gas. Extracted He was spiked with 3 He, purified using gettering methods, and analyzed on a quadrupole mass spectrometer. All degassed apatites were then retrieved, spiked with a 235 U-230 Th-145 Nd-51 V tracer, and dissolved in HNO 3 at ~90 °C for 1 hour. 238 U, 232 Th, 147 Sm and the tracer isotope analyses were acquired on an Agilent 7500 Series ICPMS for the apatites degassed at Caltech, and on a Thermo-Finnigan Element ICPMS at the University of California Santa Cruz for the apatites

Flow in the sublithospheric mantle is increasingly invoked as a mechanism to explain both modern and past surface topography, but the importance of this phenomenon and its influence at different localities are debated. Southern Africa is an elevated continental shield proposed to represent dynamically supported topography. However, this region is also characterized by a complex lithospheric architecture variably affected by Cretaceous heating, thinning, and metasomatic alteration. We used apatite (U-Th)/He thermochronometry on 15 Cretaceous kimberlites from an 600 km long transect across the Kaapvaal Craton, combined with information from xenoliths in these pipes, to determine the plateau interior erosion history. The goal was to determine the relationships with lithospheric modification patterns and thereby better isolate the sublithospheric contribution to elevation. The results document a wave of erosion from west to east across the craton from 120 to <60 Ma, initially focused along paleorivers and then retreating as a scarp across the landscape. This spatially variable erosion event was associated with limited modification of the Archean cratonic lithospheric mantle as recorded by mantle xenoliths and xenocrysts, implying that dynamic buoyancy sources may be required to explain the elevations. In contrast, off-craton to the southwest, a more pronounced regional erosion phase at 110-90 Ma was coincident with significant modification of the Proterozoic lithospheric mantle. This relationship suggests that lithospheric processes were more important contributors to erosion and topographic change off-craton than on-craton. Together, these results suggest that lithospheric architecture can have an important control on the surface expression of mantle dynamics.

Kimberlites provide rich information about the composition and evolution of cratonic lithosphere. Accurate geochronology of these eruptions is key for discerning spatiotemporal trends in lithospheric evolution, but kimberlites can sometimes be difficult to date with available methods. We explored whether (U-Th)/He dating of zircon and perovskite can serve as reliable techniques for determining kimberlite emplacement ages. We obtained zircon and/or perovskite (U-Th)/He (ZHe, PHe) dates from 16 southern African kimberlites. Most samples with abundant zircon yielded reproducible ZHe dates ( 15% dispersion) that are in good agreement with published eruption ages. The majority of dated zircons were xenocrystic. Zircons with reproducible dates were fully reset during eruption or resided at temperatures above the ZHe closure temperature prior to entrainment in the kimberlite magma. Not dating hazy and radiation damaged grains can help avoid anomalous results for more shallowly sourced zircons that underwent incomplete damage annealing and/or partial He loss during the eruptive process. All seven kimberlites dated with PHe yielded reproducible ( 15% dispersion) and reasonable results. We conducted two preliminary perovskite 4 He diffusion experiments, which suggest a PHe closure temperature of >3008C. Perovskite in kimberlites is unlikely to be xenocrystic and its relatively high temperature sensitivity suggests that PHe dates will typically record emplacement rather than postemplacement processes. ZHe and PHe geochronology can effectively date kimberlite emplacement and provide useful complements to existing techniques.

The southern African Plateau is a dominant feature of African topography reaching heights of >3,000 m, with an average elevation of 1,000 m in the predominantly low relief plateau interior. The elevated margins of the plateau drop through higher relief regions to the coastal plain (Figure 1). However, extensive debate remains regarding when and how it formed. The long wavelength topographic high in absence of collisional tectonism combined with Cretaceous kimberlite activity and a large low shear velocity province (LLSVP) in the deep mantle below southern Africa have led many to suggest uplift related to mantle processes. Given that surface uplift may be related to LLSVP development (e.g., Lithgow-Bertelloni & Silver, 1998), better