Mantle-derived magmas are among the best sources of information about the chemistry of the Earth's mantle. For the past two decades, geochemists have used ratios of certain trace elements and isotopes of these magmas, considered to be largely unaffected by crystal-liquid separation, to argue that the mantle is compositionally heterogeneous. This development has led to models of crust-mantle evolution that involve dynamic processes such as recycling of oceanic lithosphere.
Hard data supporting mantle heterogeneity come from ocean island and mid-ocean ridge basalts and are rapidly accumulating, particularly as our analytical capabilities improve. These high quality data allow us to rigorously test a fundamental assumption of geochemical mantle models: that oceanic basalts faithfully reflect particular geochemical characteristics of their mantle sources. Recent work suggests that interaction between ocean island magmas and ocean crustal material may be more common than previously thought and that such interaction may compromise chemical characteristics that were assumed to represent the mantle.