Reading garnet's signature

• Published in: Nature (London) Vol. 384; no. 6606; pp. 215 - 217
• Main Author: Hirschmann, Marc
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group 21.11.1996
• Subjects: Marine
• ISSN: 0028-0836; 1476-4687
• DOI: 10.1038/384215a0

In the quest to understand the genesis of oceanic crust, a central observation is that crust formation is not equally vigorous everywhere. In Iceland, where the mid-ocean ridge emerges above sea level, new crust is more than 15 km thick. At the other extreme, some ridges lie beneath 5,000 m of water or more, and have crust less than 4 km thick. Is this variation in crust thickness caused primarily by variation in the temperature of upwelling mantle, or by varying extents of cooling from the surface? And what effects might differences in mantle composition have? Bourdon et al. report the first worldwide compilation of ( super(230)Th)/( super(238)U) ratios for MORB (the parentheses signify abundance multiplied by the nuclide's decay constant). Garnet is the key here--basalts with ( super(230)Th)/( super(238)U) greater than unity require mantle sources that contain garnet, which is stable only in the deeper portions of MORB source regions. Bourdon et al. find that the magnitude of regionally averaged ( super(230)Th)/( super(238)U) is negatively correlated with water depth. This is best explained if the initial depth of melting is greatest for shallow ridges, allowing garnet-melt interaction over a larger part of the melting and melt-transport regime (although this interpretation requires a number of assumptions). As shown, a logical interpretation is the mantle beneath shallow ridges is hotter than that beneath deep ridges, although the data do not rule out some influence from variations in mantle composition.