Systematics of three-component, pseudo-binary mixing lines in 2D isotope ratio space representations and implications for mantle plume–ridge interaction

Abstract

Linear arrays in 2D isotope ratio space representation can be the result of mixing between three components and are referred to as pseudo-binary mixing arrays. The requirements for forming such arrays are (1) the mass fraction of at least two of the three endmembers must co-vary linearly, and (2) the elemental enrichment of the different isotopes in a given endmember relative to the third endmember must be the same (i.e., Pb 1/Pb 3 = Sr 1/Sr 3 = Nd 1/Nd 3, or equal Pb/Sr = Nd/Sr = Pb/Nd in a given endmember). Pseudo-binary mixing systematics do not require that the mixing line points to any of the three mixing endmembers. Only in the special case where the ratio of the mass fraction of two of the components is constant (e.g., the mass fraction of component 2 is always half the mass fraction of component 1) does a pseudo-binary mixing array point to a mixing endmember. Three case studies are presented which illustrate the variety of geological settings in which pseudo-binary mixing may occur. (1) Along the Mid-Atlantic Ridge (MAR), basalts between 48.5–49°S form a pseudo-binary mixing array in 2D isotope ratio space representation resulting from simultaneous mixing of the off-ridge Discovery mantle plume, ambient MORB source mantle, and a LOMU passive mantle heterogeneity thought to be delaminated subcontinental lithospheric mantle. The ratio of the mass fraction of the Discovery plume component and the LOMU component is constant, and the mixing array points towards the composition of the ambient upper mantle. (2) LREE depleted basalts from the MAR between 20–40°S form a pseudo-binary mixing array in 2D isotope ratio space representation resulting from simultaneous mixing of dispersed St. Helena plume head material, dispersed Tristan plume head material, and depleted MORB source mantle. These basalts also form smooth along ridge isotope ratio gradients. (3) Indian Ocean MORBs form a distinct array in 2D isotope ratio space representation which can be modeled using three component, pseudo ­binary mixing model. Suitable mixing endmembers are depleted MORB mantle (North Atlantic type), mildly high 206Pb/ 204Pb plume material (e.g. ‘C’ component [Hanan, B.B., Graham, D.W., 1996. Lead and helium isotope evidence from oceanic basalts for a common deep source of mantle plumes. Science 272, 991–995]), and a LOMU component representing delaminated, subcontinental lithospheric mantle dispersed into the upper mantle during the breakup of Gondwana and subsequent opening of the Indian and South Atlantic oceans. However, such a model requires single-stage mixing of the three endmembers which is not supported by the large variations in the mass fractions of the DM and plume components. Consequently, a two-stage, three-component mixing model similar to the model proposed by Hamelin et al. [Hamelin, B., Dupre, B., Allègre, C.J., 1986. Pb–Sr–Nd isotopic data of Indian Ocean ridges: new evidence of large-scale mapping of mantle heterogeneities. Earth Planet. Sci. Lett. 76, 288–298] is thought to be a more likely model for the origin of the Indian MORB array. The important difference between our model and that of Hamelin et al. is the origin of the low 206Pb/ 204Pb component. We invoke a LOMU component with an origin in the subcontinental lithospheric mantle, whereas Hamelin et al. (1986) invoke a low 206Pb/ 204Pb component representing recycled oceanic crust and sediment.