Tracking S-type granite from source to emplacement: Clues from garnet in the Cape Granite Suite

Abstract

This study investigates, via a pseudosection approach, the conditions of formation of garnet in the leucogranitic to granodioritic S-type Cape Granite Suite (CGS), South Africa. Previous work has stressed the importance of peritectic garnet entrained from the anatectic source in the petrogenesis of these granites. In this study, garnet from S-type granites of the CGS, showing as little evidence for replacement as possible, was studied for major and trace element geochemistry. Surprisingly, the compositions of all the crystals investigated are essentially identical, despite significant differences in the composition of the host granite. The garnet major element compositions are characterised by homogeneous, unzoned core domains with a relatively Mg-rich composition (Alm 69–71Py 14–21Gro 3Sps 3–5) surrounded by a more Mn-rich rim, some 200 µm wide (Alm 70–76Py 5–12Gro 3Sps 6–12). Trace element compositions are similarly characterised by unzoned cores surrounded by thin rims of relative REE enrichment. Pseudosections calculated for compositions ranging from granite to granodiorite illustrate that garnet is a stable phase in all compositions at high temperatures. Garnet core compositions equilibrated under P– T conditions of 4 to 6.2 kbars and 740 to 760 °C, whilst the rims record conditions of 2.5 to 5 kbars and 690 to 730 °C. Rare granulite-facies metamafic xenoliths also may record the conditions in the source of the granite magma and provide estimated P– T condition above 10 ± 2 kbars and 810 ± 54 °C. This estimate overlaps with the P– T conditions required for fluid-absent biotite melting, the process believed to have produced the CGS magmas within the lower crust. The pseudosections show that garnet was present in the CGS magmas from the source down to near-solidus conditions, but that the composition of peritectic garnet entrained within the source is not preserved in the magma. Calculation of the time required to homogenise garnet compositions within the magma indicates that this cannot occur by diffusion within the garnet crystals, as this would require several orders of magnitude longer than the typical duration of felsic magmatic events. Thus, the findings of this study argue for 1) entrainment of peritectic garnet into melt at the source, 2) the subsequent re-equilibration of this garnet to lower pressure and temperature conditions within the magmatic environment through a dissolution precipitation mechanism, and 3) a near-solidus complete replacement of garnet in some compositions. Collectively, these three processes explain the chemical connectedness between granites and their sources, as well as why the details of the connection have remained so elusive.