White Paper Title: 
Length scales of mantle heterogeneities and mantle control on ridge segmentation?

Isotope systematics of basalts provide information on the distribution of different mantle components as well as the length scale of mantle heterogeneity.  However, many data points are required to obtain a picture with a useful resolution.  Here we present new hafnium isotope data on over four hundred samples.  This study makes a significant step determination of both location of mantle components and order of mixing.  Our new data for mid-ocean ridge basalts (MORB) shows that on a “local” scale hafnium and neodymium isotopes are correlated, and on a global scale they form parallel arrays in this isotope space (see Fig.1).  These data are consistent with the asthenosphere consisting of domains with dimensions of several hundred to a thousand kilometers at which the fractions of the recycled materials in the mantle components are constant.  Both oceanic crust and mantle residual after extraction of MORB are recognized as recycled components and significant amounts of these components reside in the MORB-source; i.e. the convecting mantle.  The “local”, smaller, scale variation in neodymium and hafnium isotopic compositions results from the addition of a mantle component with less radiogenic neodymium and hafnium isotopic composition.  Beyond this length scale the domains differ in the amount of recycled components in the asthenospheric mantle. 

Both MORB and OIB show characteristics that indicate that the Pacific Ocean mantle is distinct from the Indian or Atlantic Ocean and is on average less depleted.  Global comparisons of ridge basalts should be cognizant of these large-scale differences.

On a smaller scale, a study along the EPR between 60N and 180N shows that the ridge segmentation coincides with chemical discontinuities.  Along this section of the EPR there are four transform fault zones and four overlapping spreading centers.  Across each discontinuity the source components change (Fig. 2).  These systematics indicate that there is no transfer of magma across the discontinuities as the chemical gradients are largest across the discontinuities.  A detailed study of the Siqueiros Fracture Zone, which has active volcanism in the fracture zone, confirms this observation. The trace element and isotopic variations of the fracture zone basalts and its adjacent segments shows that the fracture zone draws melts that can be sourced from either side of the fracture zone, but the basalts from the adjacent segments do not have a common source.

It is therefore important that the each location is placed in its regional context.  Since segmentation and source composition are related, comparison of ridge segments, especially across ocean basins should consider regional variations as well as differences in source compositions.  Since the variations in source composition can be related to the fractions of enriched and depleted components in the source, we can use the isotopic differences to estimate melt behavior.  Although differences in spreading rate have an important influence on the morphology of the ridge and the range of variations in the basalts chemistry, the fact that segmentation can be related to mantle source indicates that source composition also has a significant influence.  For example, the existence of an axial magma chamber is dependent on the magma supply rate.  Magma supply is dependent on the location of the mantle solidus which, on average, will be different for the Atlantic Ocean basin which has a higher fraction of a depleted component than for the Pacific Ocean basin.