HR: 0830h
AN: V11E-0532    [PDF]
TI: Using Breccia-Hosted Spinels of Abyssal Peridotites to Obtain a Representative Local Mantle Composition
AU: * Hellebrand, E
EM: ehelle@mpch-mainz.mpg.de
AF: Max-Planck-Institut fuer Chemie, Postfach 3060, Mainz, 55020 Germany
AU: Snow, J E
EM: jesnow@mpch-mainz.mpg.de
AF: Max-Planck-Institut fuer Chemie, Postfach 3060, Mainz, 55020 Germany
AU: Emami, M
EM: emami@mpch-mainz.mpg.de
AF: Max-Planck-Institut fuer Chemie, Postfach 3060, Mainz, 55020 Germany
AU: Dick, H J
EM: hdick@whoi.edu
AF: Woods Hole, Oceanographic Institution, Woods Hole, MA 02543 United States
AB: Surprisingly little is known about the extent, scale, and causes of local (outcrop/dredge) heterogeneity in the oceanic mantle. Our primary sources of information are abyssal peridotites, which are fragments of the residual MORB mantle tectonically exposed on the ocean floor at mid-ocean ridges. On a ridge scale it is impossible to objectively pick representative samples for detailed petrological and geochemical analyses, due to analytical limitations and financial restrictions, as well as operator-biased sample selection. This is particularly important for the ultraslow spreading Gakkel Ridge (Arctic Ocean), where thousands of individual peridotite specimens from nearly forty sampling locations are available. Preliminary conventional (i.e. major and trace element mineral analyses in-situ) petrological investigations at Gakkel Ridge have revealed variable dredge-scale heterogeneities, which are related to regional changes in the extent and relative proportion of melting/melt-migration. In order to better assess the extent and distribution of dredge-scale chemical variations, we have separated spinel grains out of carbonate-cemented peridotite breccias that were collected along with normal serpentinized peridotites (and which are common on the ocean floor). The rationale for this is that the spinel clasts may provide a more representative composition of the local outcrop, or slope from which the serpentinite fragments were sedimented by mass wasting. In a pilot study, we selected 29 breccia samples from 7 dredge hauls. For each of these dredge hauls major and trace element mineral data of more than 8 normal residual abyssal peridotite (RAP) samples are available. The major element spinel compositions of the RAP hand specimens (n=95) were then compared with those obtained on breccia-hosted (BH) spinels (n=1300) from the same dredge haul. The agreement between BH spinels and those from the RAP is very good. As expected, BH-spinels cover a larger range than RAP-spinels, although each individual breccia only covers a part of the entire spectrum within a single dredge. This may mean that the spinels are derived quite locally and not mixed much. The Cr\#-Ti systematics of the spinels can be used to infer degrees of melting, extent of reaction with percolating melts, and spatial distribution and abundance of dunites channels. Besides spinels from residual peridotites, the breccias host spinels from peridotites affected by high-level melt percolation, such as plagioclase-bearing peridotites or ones affected by crosscutting dykelets. In most cases, it possible to distinguish the non-residual from the residual ones (e.g. dredge D34: 40% residual, 50% dunitic, 10% vein-influenced; n=162). Despite some limitations (higher spinel mode in dunite, possible preferential weathering of dunite), this may prove a useful tool to estimate the local proportions of residual mantle, melt transport channels, and the volume of mantle affected by late-stage crosscutting dykelets.
DE: 1025 Composition of the mantle
DE: 3035 Midocean ridge processes
DE: 3640 Igneous petrology
DE: 3655 Major element composition
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting