Answers are grounded in this paper and cite their sources.
Ultramafic intrusions in the shallow oceanic crust are rarely observed, limiting our understanding of their formation and evolution. Macquarie Island, located between the Australian and Pacific plates, southwest of New Zealand, is a subaerial section of in-situ oceanic crust. North Mountain on the island hosts an ultramafic intrusion at the gabbro-dike boundary. The intrusion formed after the initial crust-forming event, following the development of an off-axis oceanic core complex and late, off-axis magmatism. We test whether the intrusion samples formed from a single melt or if they record a reactive overprint. Although previously classified as plagioclase-bearing wehrlite, we show that our samples are more similar to olivine-rich troctolites commonly reported from the oceanic crust-mantle boundary. Mineral modal abundances and mineral chemistry indicate an origin through melt-rock reactions between a MORB-type melt and a pre-existing olivine-rich matrix which was part of the uplifted mantle or lower crustal sequence of the oceanic core complex. The impregnating melt dissolved the precursor mantle minerals, and crystallized interstitial clinopyroxene and plagioclase, forming an olivine-rich troctolite. Contrary to other global olivine-rich troctolites, the intrusion was surrounded by cooled and solidified shallow crust instead of an unsolidified gabbroic crystal mush, entrapping the evolved interstitial melts into the crystallizing troctolite body. We show that the abundant magmatic amphibole crystallized from these highly evolved, late-stage interstitial melts. Heat loss to the surrounding crust resulted in rapid cooling of the intrusion that is recorded by the high CaO contents preserved in olivine surrounded by clinopyroxene. The North Mountain intrusion samples reveal that olivine-rich troctolites analogous to the olivine-rich troctolites of the crust-mantle transition zone can form off-axis at shallow crustal levels following the uplift of mantle rocks.
No takes yet. Share an insight, caveat, or question.
Kolehmainen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: