There are clear indications of very ancient sub-continental lithospheric mantle (SCLM). Samples are well known from the mid-Archaean and some may be > 3.8 Ga old. Understanding the origin of the SCLM is essential to understanding Archaean crust formation, not least because it is the principal reason why the most ancient Archaean crust is preserved. However, the relationship between the Archaean SCLM and Archaean continental crust is an enigma. Firstly, Archaean SCLM is particularly thick and has a very depleted composition relative to more recent SCLM. Secondly, whilst there is a very clear temporal relationship between Archaean crust and lithosphere, indicating that they are part of a common process, the melt-restite relationship, is obscure. Models have been proposed for the extraction of the continental crust from the SCLM, and more believably, the extraction of a komatiitic melt, although neither model satisfies mass balance constraints.

A recent version of the komatiite extraction model was proposed by Parman et al. (South African J. of Geology, 107, 107-118, 2004) suggesting that the Barberton komatiites were extracted from the SCLM beneath the Kaapvaal Craton and are the compliment to depleted Kaapvaal peridotitic lithosphere. An attractive feature of this model is that it views SCLM formation in the context of subduction and the modification of the mantle wedge. Unfortunately, again, it does not satisfy mass balance arguments.

Here an alternative model for the origin of the SCLM is proposed. This model starts with the premise of crust-formation through slab-melting, and that as a result of this process the mantle wedge becomes impregnated with a low % of TTG melt. If this, enriched primitive mantle, is chosen as the starting composition for the crust-SCLM system, then the removal of about 30% basaltic melt (+/- a small fraction of komatiite) will leave a residue with the composition of depleted Archaean SCLM. The advantages of this model are that (a) the removal of a more modest fraction of basalt, rather than a large fraction of komatiite, is more consistent with what is observed in Archaean greenstone belts, and (b) it provides an integrated solution to the SCLM problem, in that there is an explanation for the temporal link between SCLM and Archaean crust, whilst the lack of a direct genetic relationship is maintained.