Lanthanide abundances basalts

The lanthanide abundance pattern for Shergotty (table 9, fig. 7) is unique among meteorites and indicates a complex prehistory in the Martian mantle from which they appear to have been derived, as basalts, by partial melting. The enriched pattern of the heavy lanthanides (Gd-Lu) resembles that of pyroxenes (the parent rocks appear to have been pyroxene cumulates). It provides no evidence that garnet was a residual phase in the source from which these basalts were derived, for, if so, the reciprocal pattern would be displayed. Leaching experiments show that most of the lanthanides are contained in accessory phases (whitlockite and apatite) rather than in the major mineral phases. 

Fig. 9. Lanthanide abundance patterns for high-Ti, high-K and high-Al lunar basalts, again showing the general depletion in Eu, which reflects the pattern in the source regions at depths of 150-400km within the moon. (Data are from table 11.)...

Fig. IZ Simplified internal structure of the moon, showing the mineralogically zoned source regions from which the mare basalts were derived, the feldspar-rich crust, and KREEP. Lanthanide abundance patterns for these various regions are depicted on the right with the approximate concentrations relative to average chondrites indicated. The lunar interior is undoubtedly more complex both vertically and laterally than depicted.

Fig. 23. Lanthanide abundance patterns for Mid-Ocean Ridge Basalts (MORE) and Ocean-Island Basalts (OIB) (data are from table 16). Note the depletion in the light lanthanides (La-Sm) in MORE derived from a depleted mtmtle source, and the enriehment of light lanthanides in OIB. Nd isotopic evidence indicates that the source of both these rock types was characterized by long term depletion of Nd relative to Sm, indicating the observed light lanthanide enrichment in OIB is a recent event.

The close similarity of the lanthanide distributions in ocean floor volcanics to that of the chondrites is further evidence that Earth has the same overall average relative lanthanide abundances as the chondrites. Otherwise, the uniformity of these melt products of the mantle found in all the oceans of the world would seem to be fortuitous. The distribution is not unmodified from that of the chondrites, and the lavas are not primitive or first-generation melting products of a primitive terrestrial mantle. Variations in lanthanide concentrations and relative abundances among ocean floor basalts are mainly the result of minor inhomogeneities in the mantle source regions, small differences in conditions of partial melting, and crystal fractionation of the lavas prior to eruption. 

Interpreting the lanthanide distributions (and other information) in volcanics from even a single volcano can illustrate the complexity of processes operating at a continental boundary. Condie and Hayslip (1975) have studied lanthanides in the younger lavas of the Medicine Lake, California shield volcano. This volcano first formed as much as half a million years ago and produced abundant basaltic andesite and andesite lavas, the youngest of which are least 10 thousand years old. Then relatively small amounts of more acidic (siliceous) lavas were produced. 

There are substantial difficulties with this explanation. The island arc volcanics in question are relatively deficient in light lanthanides compared with the NASC and the Precambrian sediments. No combination of their distribution and that of Eu-deficient crustal material can produce the NASC-like distribution with increased Eu. Also, the sediments showing Eu anomalies include well differentiated shales, sands, and carbonates, probably not of eugeosynclinal origin. Finally, several of the Precambrian sediments had relative Eu abundances greater than that of the chondrites and, therefore, the island arc basalts. 

Basalts, basaltic andesites, and andesites with this distribution are common in some island arcs (e.g., Jakes and Gill, 1970 Ewart et al., 1973 Taylor et al., 1969). Their presence is believed to result from melting of subducted oceanic crust. By and large, the sediment layers which lie above the ocean floor tholeiites and are derived mainly from continental material are not subducted but piled up against continental margins in some manner that prevents their modifying significantly the trace element and isotopic abundances of oceanic crustal matter in the production of this class of island arc volcanics. Nor does ocean water severely modify the lanthanide distributions in volcanics that are extruded under... 

---