Dunite magma or ultramafic cumulates? A discussion of Griffin et al. “Intrusion and contamination of high-temperature dunite magma: the Nordre Bumandsfjord pluton, Seiland, Arctic Norway” Contrib. Mineral. Petrol. (2013) 165:903–930

Abstract

It was with considerable surprise that I noticed the publication in CMP of the paper by Griffin et al. (2013), the essence of which seems to have been resurrected after being shelved for almost 30 years. The paper is based on, and reiterates, the ‘‘unequivocal evidence for the intrusion of very high temperature, very fluid ultramafic (dunitic) magmas’’ originally presented by Sturt et al. (1980). These authors make the extraordinary claim that 60–70 % of the rocks in the Nordre Bumandsfjord pluton are the result of contamination of primary dunite magma on a massive scale by assimilation of gabbroic and other host rocks as well as hybridisation with anatectic melts derived from them. The existence of dunite magma is inferred from the existence of narrow dykes in gabbros and peridotites ‘‘varying in composition from dunite to feldspathic lherzolite; the thinnest veins are commonly dunite’’. Sturt et al. and Griffin et al. hence take the erroneous position that any coarse-grained igneous rock that occurs in a narrow dyke must once have been a fluid magma of the same composition. The aim of this contribution is to bring to your reader’s attention alternative interpretations based on exhaustive fieldwork over several years by Mike Bennett (now sadly deceased), myself and our students in all of the larger ultramafic complexes within the Seiland Igneous Province in northern Norway. In a 40-page-long publication, Bennett et al. (1986) summarised the lithologies and structural relationships within the Melkvann and Reinfjord complexes that we had studied in most detail, as well as our observations in the Kvalfjord and Bumannsfjord plutons. According to Griffin et al., the Melkvann and Kvalfjord complexes have only been described briefly, but I would refer readers to the publications by Yeo (1984), Leaver et al. (1989), Leaver (1990) and Svensen (1990) (none of which are referred to by Griffin et al.) for more detailed petrological documentation of these particular complexes, as well as to Emblin’s (1985) comprehensive account of the Reinfjord ultramafic complex. Bennett et al. concur with Griffin et al. that there is overwhelming field evidence in the Seiland-Stjernoy ultramafic complexes for both in situ partial melting and assimilation of gabbroic rocks; contaminated igneous rocks enclosing abundant xenoliths and xenocrysts are obvious and widespread (the most extreme varieties are olivine melagabbros). Even so, Griffin et al. encountered some difficulty in explaining their analysed samples of peridotite simply as binary mixtures of their inferred dunite magma and sampled gabbroic rocks and were forced to appeal to the extraction of a component selectively enriched in Carich pyroxene from the host gabbros. The possibility of the presence of cumulus or intercumulus pyroxene in the peridotites compared to the dunites was not explored. As pointed out by Bennett et al., assimilation of gabbroic rocks by some presumed dunite melt or mixing with an anatectic melt can hardly account for olivine clinopyroxenite that is volumetrically the most important ultramafic rock type in all of these complexes. Bennett et al. regarded the majority of the ultramafic rocks of the Melkvann, Nordre Bumannsfjord and Kvalfjord complexes as igneous cumulates, even those forming narrow, dilational dykes. One of the principal reasons for Communicated by J. L. R. Touret.