Multiple Diamond Sources at the Kalahari Craton Margin: Letseng-la-terai, Lesotho

Paula McDade Dept. of Geology and Applied Geology, University of Glasgow, Glasgow G12 8QQ, Scotland, UK; present address: Dept. of Geology and Geophysics, University of Edinburgh, Edinburgh EH9 3JW, UK

Jeff W. Harris Dept. of Geology and Applied Geology, University of Glasgow, Glasgow G12 8QQ, Scotland, UK

Syngenetic mineral inclusions preserve a record of the chemical environment at the time of formation of the host diamond. This study evaluates the syngenetic inclusion chemistry and carbon isotope signatures of 48 diamonds collected from the Letseng-la-terai kimberlite pipe, Lesotho.

Diamond parageneses represent different chemical environments of formation and are named according to the mantle lithology the inclusion chemistries reflect. The variation in the Cr2O3 vs CaO contents of the garnet inclusions from Letseng reveal four distinct populations; (i) lherzolitic (1) (ii) CaO poor harzburgitic (16) (iii) CaO rich eclogitic (2) and (iv) websteritic (Cr2O3<2wt%, variable CaO) (2). Orthopyroxene and olivine analyses were typical of either the lherzolitic or harzburgitic parageneses (opx Mg# ranging from 0.94 to 0.96, olivine Mg# from 0.925 to 0.950). One omphacitic clinopyroxene from the eclogitic parageneses was recovered, and a second co-existed with a garnet which was websteritic. All these analyses fall within the range of mineral inclusions in diamonds worldwide, but are more magnesium-rich than the majority.

Some less common inclusions recovered were eclogitic coesites (4), a sulphide (pyrite exsolving from a monosulphide solid solution), three calcites (one co-existing with olivine), an octahedral diamond, and a single ferropericlase. The latter four inclusion types are of uncertain paragenesis.

Calcite inclusions in diamond are typically of epigenetic origin, formed by CO2-rich fluids infiltrating the crystal through fractures and altering pre-existing inclusions. The absence of observable fracture systems around the carbonates, which would have allowed fluid access, suggests that in this case all the inclusions are syngenetic.

Ferropericlase and perovskite structured enstatite are the predicted breakdown products of g spinel structured olivine (Ringwood, 1966). The presence of a ferropericlase inclusion suggests that this diamond was formed in the lower mantle at depths greater than 670km, but the absence of a co-existing MgSiO3 phase renders the paragenesis of this diamond inconclusive. It may have formed under upper mantle conditions if the oxygen fugacity and silica activity had been sufficiently low.

Geothermobarometry calculations suggest inclusion formation and subsequent diamond growth at temperatures between 1125ƒC and 1323ƒC, and pressures between 52kb and 68kb.

The d13C stable isotope ratios of the diamonds were measured, revealing at least two significantly different populations. The majority of the data fall within the range -2.1” to -11.24”, while a smaller cluster extends from -15.79” to -16.49” (relative to PDB).

It is believed that the variation in fractionation of carbon during diamond growth, within a single reservoir, will not exceed 5” (Kirkley et al., 1991). Applying this finding to Letseng and combining the d13C data with the mineral chemistry of the inclusions, indicates that this diamond suite represents a kimberlite sampling event from at least seven distinct mantle temporal sources. The results of this study show significant similarities to the findings of Deines et al. (1991) who report that the neighbouring Jagersfontein kimberlite sampled at least six distinct diamond source regions. This suggests that the mantle beneath the south-eastern margin of the Kapvaal craton is compositionally extremely heterogeneous, relative to the mantle beneath an intra-cratonic locality.


Deines, P., Harris, J.W. & Gurney, J.J., Geochim. cosmochim. Acta. 32, 2615-2625 (1991).

Kirkley, M.B., Gurney, J.J., Otter, M.L., Hill, S.J. & Daniels, L.R., Applied Geochem. 6, 477-494 (1991).

Ringwood, A.E., In Advances in Earth Science (ed. Hurley, P.M.) 357-399 (Cambridge, Mass., M.I.T. Press, 1966).