Discussion of “Kimberlites and aillikites as probes of the continental lithospheric mantle”, by D. Francis and M. Patterson (Lithos v. 109, p. 72–80)

The origin of kimberlites and volatile-rich alkaline ultramafic magmas such as aillikites has been a matter of considerable scientific interest and the recent paper by Francis and Patterson (Francis and Patterson, 2009. Kimberlites and Aillikites as probes of the continental lithospheric mantle. Lithos 109, 72–80) attempts to summarize advances in this research field. They introduce new major element geochemical discrimination diagrams and present a petrogenetic model that allegedly has implications for diamond exploration. Contrary to the statements of Francis and Patterson we argue that kimberlite bulk chemical compositions are not determined by contamination with lithospheric mantle material alone and cannot be used to predict reliably the diamond grade of new discoveries. We dispute their petrogenetic model, specifically the proposal that calls for carbonate-rich aillikite melts being parental to kimberlite magmas. Rather, we promote the idea of carbonated ultramafic lamprophyre magmas, such as aillikites, being parental to many carbonatite intrusions in areas of rifted cratonic lithosphere, from which archetypal kimberlites appear to be absent. Furthermore, we believe that the geochemical discrimination diagrams (Si versus K and Fe, respectively) introduced by Francis and Patterson do not effectively distinguish between kimberlites and aillikites because there is significant data overlap and bias. Of particular concern is the omission of relevant compositional data for the ultramafic lamprophyre group. In this discussion we re-iterate that bulk chemical analyses of kimberlites and alkaline ultramafic rocks need to be accompanied by petrographic descriptions and mineralogical data. Only by the inclusion of such data can the petrogenesis of the rocks considered be discussed.