“…Additionally, their particularly low TiO 2 and HFSE abundances, their locally high Al 2 O 3 /TiO 2 ratios (higher than primitive mantle values of ∼21) and marked depletion in Nd isotopic composition (ε Nd (t = 2.54) value up to 4.45), indicate that their source region is highly depleted and refractory peridotitic mantle that has likely experienced one or more episodes of basaltic melt extraction prior to remelting (e.g., Sun and Nesbitt, 1978;Cameron et al, 1983;Crawford et al, 1989). On the other hand, their siliceous, strong LILE and LREE enrichment and HFSE depletion may be ascribed to melting of metasomatically enriched lithospheric mantle (Weaver and Tarney, 1981;Fisk, 1986;Falloon and Danyushevsky, 2000;Smithies et al, 2004a). Hence, their petrogenesis is most likely comparable to that of typical Phanerozoic and Archean boninites (e.g., Crawford et al, 1989;Falloon and Danyushevsky, 2000;Smithies et al, 2004a), being especially similar to the Late Paleocene boninites from the Cape Vogel peninsula of Papua New Guinea (i.e., PNG boninites; König et al, 2010) and the Mesoarchean Whundo boninite-like rocks from the Pilbara Craton in northwestern Australia as shown in Fig.…”