PALEOMAGNETISM OF MESOPROTEROZOIC ALKALINE AND CARBONATITIC COMPLEXES IN THE KAAPVAAL CRATON, SOUTH AFRICA

The Kaapvaal craton in South Africa is penetrated by a number of ~1.4–1.35 Ga alkaline and carbonatitic complexes scattered over a wide area in the central part of the craton. The complexes tend to occur near or along Archean terrane boundaries or younger extensional faults, implying that zones of lithospheric weakness controlled ascent of mantle-derived melts during intraplate magmatism. Trachytic to phonolitic volcanic vents and associated lava/pyroclastic sequences are partly preserved in some complexes, and shallow-level intrusive rocks occur in varying proportions (e.g., syenite, nepheline syenite, pyroxenite, ijolite, carbonatite). SHRIMP U-Pb zircon or titanite crystallization ages obtained from several of the complexes range from 1397 ± 47 to 1374 ± 10 Ma.

Because of their intraplate setting, the alkaline complexes potentially represent ideal targets for paleomagnetic studies, particularly because there are no robust paleomagnetic data for the Kaapvaal craton in this part of the Mesoproterozoic. We have carried out reconnaissance sampling in three of the main complexes. Four widely spaced sites were sampled in nepheline syenites along the margins of the ~1.4 Ga Pilanesberg Complex, which occupies an area of ~530 km2 and is the largest of the complexes. ~150 km to the east, three sites were sampled in nepheline syenite, alkali-feldspar syenite, and trachytic lava within two igneous centers (Franspoort and Leeuwfontein) in the closely grouped, 1.4–1.37 Ga Pienaars River Complexes. Still farther east, three sites were sampled in dolomite carbonatite, nepheline syenite, and ijolite in the Spitskop Complex, which has yielded an Rb-Sr isochron date of 1341 ± 37 Ma.

All samples were subjected to progressive thermal demagnetization in ten or more steps. Four sites did not yield usable data. The remaining six sites (two from each complex) had statistically acceptable data, with alpha 95 values as low as 3°. However, their pole positions are widely scattered, including sites within a single complex, and the scatter is much greater than could be accounted for by undetected structural tilts. The divergence of poles also precludes a common remagnetization event. Detailed rock magnetic analyses may salvage some of the data, but the wide scatter in pole positions presently remains unexplained.