Characteristics and Origin of the Mantle Root Beneath the Murowa Diamond Mine: Implications for Craton and Diamond Formation

Abstract

Abstract The Murowa and Sese kimberlites erupted through the southern margin of the Zimbabwean craton. These kimberlites provide a unique sample of the continental lithospheric mantle in that area through their entrained mantle xenolith and xenocryst cargo. Mantle xenoliths have only been obtained from the Murowa locality so far and thus they form the focus of this review. Ultradepleted chromite-harzburgite and chromite-dunite rock units dominate the Murowa mantle xenolith inventory. No eclogite xenoliths have been found and eclogitic garnets are extremely scarce in the minerals analyzed from heavy mineral concentrate. The very low bulk rock Al and Ca contents of the Murowa peridotites, along with their extreme Pt and Pd depletions, require extensive melt extraction—to in excess of 40% melting. At the same time, their high bulk rock Cr# (100☼ Cr/(Cr+Al)) indicates that melting took place at relatively low pressures (<<5 GPa). Such high bulk rock Cr#s (median = 0.60) are considerably higher than those of peridotites from the nearby Venetia mine (median = 0.33) and have only been found elsewhere in cratonic peridotites from the North Atlantic craton (median = 0.89). Similar elevated bulk Cr# values are matched by spinel peridotites derived from Phanerozoic collision zones as ophiolites. This similarity favors a model in which this part of the cratonic lithosphere was formed by the subduction of peridotite that had undergone multistage low-pressure melt extraction, likely in an Archean mantle wedge that underwent flux-melting within a subduction zone prior to lateral compression to form nascent cratonic lithospheric mantle. Preliminary quantitative fitting of mantle geotherms derived from thermobarometry data is hampered by the scarcity of suitable clinopyroxene grains and is highly dependent on assumptions made regarding crustal heat production, especially in the lower crust. Nonetheless, at ~540 Ma, the time of kimberlite emplacement, the lithosphere beneath the southern edge of the Zimbabwe craton can be constrained to be approximately 200 km deep, slightly shallower than the 220-km depth estimated for lithosphere beneath the Venetia kimberlite, intruded through the Limpopo Complex. The presence of deep, ~200-km-thick lithospheric mantle beneath Murowa and Sese in early Cambrian times agrees with the minimum pressure estimates derived from Cr-Ca relationships in concentrate garnets. This estimate is close to that derived from surface-wave seismic studies and indicates that the thickness of the craton root beneath the southern Zimbabwe craton and the Limpopo Complex has not varied significantly in the last 500 m.y. The ultradepleted nature of the Murowa peridotites, together with the scarcity of eclogite/pyroxenite components, indicates a different petrogenetic history for the lithospheric root beneath the southern Zimbabwe craton compared with the mantle roots beneath the Limpopo Complex (Venetia) and the Kaapvaal craton to the south. The very high fraction (73%; n = 150) of low-Ca, high-Cr harzburgitic “G-10” garnets in the mantle garnet concentrate population at Murowa, along with their Cr-Ca relations, is consistent with the high diamond grade (0.7 ct/metric ton). The likely metasomatic origin for G-10 garnets along with the abundance of ultradepleted chromite-bearing peridotites in the Murowa mantle xenolith suite indicates that this lithology, if present in the lithosphere in the diamond stability field, may be a critical starting component for a variety of diamond- formation events in cratonic lithosphere.