Interpretation of trace element and isotope features of basalts: Relevance of field relations, petrology, major element data, phase equilibria, and magma chamber modeling in basalt petrogenesis

The concentrations and ratios of the major elements determine the physical properties and the phase equilibria behavior of peridotites and basalts in response to the changing energy contents of the systems. The behavior of the trace elements and isotopic features are influenced in their turn by the phase equilibria, by the physical character of the partial melting and partial crystallization processes, and by the way in which a magma interacts with its wall rocks. Concentrating on the trace element and isotope contents of basalts to the exclusion of the field relations, petrology, major element data, and phase equilibria is as improvident as slaughtering the buffalo for the sake of its tongue. The crust is a cool boundary layer and a density filter, which impedes the upward transfer of hot, dense primary picritic and komatiitic liquids. Planetary crusts are sites of large-scale contamination and extensive partial crystallization of primitive melts striving to escape to the surface. Escape of truly unmodified primitive melts to the surface is a rare event, requiring the resolution of daunting problems in chemical and mechanical engineering. Primary status for volumetrically abundant basalts such as mid-ocean ridge basalt, ocean island basalt, and continental flood basalts is denied by their low-pressure cotectic character, first remarked upon on petrological grounds in 1928 and on experimental grounds in 1962. These basalt liquids are products of crystal-liquid separation at low pressure. Primary status for these common basalts is further denied by the phase equilibria of such compositions at elevated pressures, when the required residual mantle mineralogy (magnesian olivine and orthopyroxene) is not stable at the liquidus. It is also denied by the picritic or komatiitic nature of partial melts of candidate upper-mantle compositions at high pressures-a conclusion supported by calculation of the melt composition, which would need to be extracted in order to explain the chemical variation between fertile and residual peridotite in natural ultramafic rock suites. The subtleties of magma chamber partial crystallization processes can produce an astounding array of pseudospidergrams, a small selection of which have been explored here. Major modification of the trace element geochemtistry and trace element ratios, even those of the highly incompatible elements, must always be entertained whenever the evidence suggests the possibility of partial crystallization. At one extreme, periodically recharged, periodically tapped magma chambers might undergo partial crystallization by similar to95% consolidation of a succession of small packets of the magma. Refluxing of the 5% residual melts from such a process into the main body of melt would lead to eventual discrimination between highly incompatible elements in that residual liquid comparable with that otherwise achieved by 0.1 to 0.3% liquid extraction in equilibrium partial melting. Great caution needs to be exercised in attempting the reconstruction of more primitive compositions by addition of troctolite, gabbro, and olivine to apparently primitive lava compositions. Special attention is focussed on the phase equilibria involving olivine, plagioclase (i.e., troctolite), and liquid because a high proportion of erupted basalts carry these two phases as phenocrysts, yet the equilibria are restricted to crustal pressures and are only encountered by wide ranges of basaltic compositions at pressures less than 0.5 GPa. The mere presence of plagioclase phenocrysts may be sufficient to disqualify candidate primitive magmas. Determination of the actual contributions of crustal processes to petrogenesis requires a return to detailed field, experimental, and forensic petrologic studies of individual erupted basalt flows; of a multitude of cumulate gabbros and their contacts; and of upper-mantle outcrops. Copyright (C) 2002 Elsevier Science Ltd.