Mantle peridotites of ophiolites rarely preserve reliable records of paleo-oceanic lithospheric mantle

Mantle peridotites of ophiolites have traditionally been interpreted as fragments of residual oceanic lithospheric mantle depleted by partial melting. However, recent petrological, geochemical and isotopic data suggest that this hypothesis needs to be reconsidered. Spinel grains in ophiolitic mantle peridotites, previously believed to record melt extraction, exhibit significant compositional variations that are incompatible with partial melting. These grains occur mostly as interstitial, anhedral crystals associated with clinopyroxene or as inclusions in silicates; they commonly also contain a variety of inclusions. Their elemental compositions are highly variable at both intra- and inter-grain, as well as intra- and inter-sample scales. In addition to association with spinel, interstitial clinopyroxene is also widespread in ophiolitic mantle peridotites. Individual clinopyroxene grains show positive correlations between Al₂O₃ and Cr₂O₃ contents, akin to those observed in orthopyroxene, which deviate from partial melting trends. Reactive and replacive features are well-developed in orthopyroxene grains. Olivine has anomalously variable Li isotope compositions and generally heavier Fe and Cr isotope compositions than coexisting spinel, contrary to theoretical predictions and results from mantle peridotite xenoliths. Additionally, aggregates of olivine, orthopyroxene, clinopyroxene, spinel and amphibole are also commonly observed in many samples. All of these characteristics suggest significant post-crystallization modifications of the constituent minerals in the mantle peridotites of ophiolites. Amphibole, a widespread hydrous mineral in mantle peridotites of ophiolites, serves as robust evidence of the past presence of hydrous melts/fluids. It typically exhibits variable compositions that can be distinguished from grains in other lithologies or different tectonic settings. These melts/fluids are most likely related to the intrusion of dunite, chromitite, clinopyroxenite, and wehrlite bodies, which may form layered or podiform occurrences in mantle lherzolites and harzburgites. The mantle peridotites can be interlayered with dunites at scales ranging from several centimeters to hundreds of meters in length and several centimeters to several meters in thickness. Drill cores of ophiolitic peridotites reveal a wide range of lithologies and extensive geochemical heterogeneity in the mantle sequences of investigated ophiolites. The modified geochemical signatures found in minerals and bulk-rock samples go far beyond what could be produced by partial melting alone; some peridotites may have even formed through the direct accumulation of crystals from melts. These active melts/fluids are inferred to be hydrous and enriched in Mg and Ca. They not only introduced exotic components to metasomatize the mantle peridotites but also facilitated compositional exchange between various lithologies and both chromite and silicates. The possible origins for the melts/fluids are proposed: dehydration and melting of subducting slabs involving melting of the mantle wedge, and the release of hydrous melts/fluids either from chromitite bodies or metamorphic sole rocks. As a result, the mantle peridotites of ophiolites can no longer be considered as paleo-oceanic lithospheric mantle and should not be used to identify paleo-oceans' reconstructions and to constrain global tectonic processes.