HR: 17:00h
AN: V22H-05    [PDF]
TI: Excess Olivine and Positive FeO-MgO Trend in Bulk-rock Abyssal Peridotites as a Consequence of Porous Melt Migration Beneath Ocean Ridges
AU: * Niu, Y
EM: yniu@mail.uh.edu
AF: Department of Geosciences, University of Houston, 4800 Calhoun Road, Houston, TX 77204 United States
AB: Abyssal peridotites (AP) are mantle melting residues for mid-ocean ridge basalts (MORB) [1-4]. Recent studies [5-8] have confirmed the complementary relationship between MORB and AP, but also revealed the hidden complexities in these peridotites such as olivine addition [5-7] and melt refertilization [7,9,10]. These same studies [5,6] have immediately excited serious debates on AP petrogenesis [11-15]. These debates are stimulating and healthy towards an improved understanding of ocean ridge magmatic processes. However, a straightforward interpretation [5,6] has been widely misunderstood because of the influence of [13]. A clarification is necessary. {\it Niu, Langmuir and Kinzler} [5] showed that a positive FeO-MgO trend exists in bulk-rock AP samples reconstructed using mineral modes and compositions [3,4]. Such a positive trend is inconsistent with AP being simple melting residues, but consistent with AP being melting residues plus excess olivine [5,6]. Using their {\it site averages} of reconstructed bulk-rock AP data, {\it Baker and Beckett} [13] countered that the positive FeO-MgO correlation by {\it Niu et al.} [5] is an artifact and there is no evidence for significant olivine accumulation in AP. The clarification here includes the following valid statements: (1) the {\it site averages} by [13] simply {\it cannot} be derived from their own {\it unaveraged} data; (2) the unaveraged data by [13], as expected, do show a positive FeO-MgO trend as shown by [5,6]; (3) the positive FeO-MgO trend is not an artifact as this trend is also clear in the unaveraged data of [13]; (4) excess olivine is already evident in the original modal data [3,4]; and (5) olivine addition is observed petrographically [7]. Therefore, the positive FeO-MgO trend defined by bulk-rock AP samples is characteristic of AP. Addition of olivine in AP is a consequence of cooling of ascending melts migrating through advanced residues in the "cold" thermal boundary layer beneath ocean ridges [5,6]. This is consistent with bulk-rock trace element data; excess olivine correlates with elevated abundances of incompatible elements (trapped melts) [7]. As olivine addition and incompatible element enrichments are observed on thin-section scales, AP record porous melt migration although dunite channels could be volumetrically more important in melt transport [16]. Massif and ophiolitic peridotites should differ from AP if they did not have the same histories as AP had beneath ocean ridges. References: [1] Dick et al., {\it Earth Planet. Sci. Lett.}, {\it 69}, 88-106, 1984; [2] Michael and Bonatti, {\it Earth Planet. Sci. Lett.}, {\it 73}, 91-l04, 1985; [3] Dick, {\it Geol. Soc. Lon. Spec. Pub.}, {\it 42}, 71-105, 1989; [4] Johnson et al., {\it J. Geophys. Res.}, {\it 95}, 2661-2678, 1990; [5] Niu et al., {\it Earth Planet. Sci. Lett.}, {\it 152}, 251-265, 1997; [6] Niu, {\it J. Petrol.}, {\it 38}, 1047-1074, 1997; [7] Niu and H\'{e}kinian, {\it Earth Planet. Sci. Lett.}, {\it 146}, 243-258, 1997; [8] Niu and H\'{e}kinian, {\it Nature}, {\it 385}, 326-329, 1997; [9] Elthon, {\it J. Geophys. Res.}, {\it 97}, 9015-9025, 1992; [10] Niu and Greig, {\it Eos Trans. AGU, Suppl.}, {\it 81 (48)}, F1282, 2000; [11] Walter, {\it J. Petrol.}, {\it 40}, 1187-1193, 1999; [12] Niu, {\it J. Petrol.}, {\it 40}, 1195-1203, 1999; [13] Baker and Beckett, {\it Earth Planet. Sci. Lett.}, {\it 171}, 49-61, 1999; [14] Asimow, {\it Earth Planet. Sci. Lett.}, {\it 169}, 303-319, 1999; [15] Lundstrom, {\it Nature}, {\it 403}, 527-530, 2000; [16] Kelemen et al., {\it Nature}, {\it 375}, 747-735, 1995.
DE: 3035 Midocean ridge processes
DE: 3040 Plate tectonics (8150, 8155, 8157, 8158)
DE: 3640 Igneous petrology
DE: 3655 Major element composition
DE: 8434 Magma migration
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting