ND ISOTOPES AS A TEST FOR THE EPISODIC GROWTH OF CONTINENTAL CRUST

The interpretation of zircon age peaks continues to be controversial: do they represent increased growth rate or increased preservation rate of continental crust? To further investigate this issue, we use epsNd(t) of granitoids from our growing database (currently 5760 samples). We have used an arc mantle source (Nd = 10 ppm) with an assumed epsNd(t) of +8 radiating from an epsNd(t) of zero at 4.4 Ga. We assume mixing between the arc mantle and a continental source (Nd = 32 ppm) whose epsNd(t) is defined by the compositions of the most negative epsNd(t) granitoids in the data set. Using this procedure we calculate that granitoids with epsNd(t) ~ 0 contain ca. 90% arc mantle. We define juvenile input as epsNd(t) ≥ 90% of the arc mantle and reworked as <90% of this source. EpsNd(t) age peaks for both juvenile and reworked continental crust coincide with zircon age peaks at 3300, 3000, 2700, 2500, 2100, 1900, 1760, 1650, 600 and 300 Ma. However, there is no epsNd(t) peak at 1100 Ma corresponding to a major zircon age peak of this age. Zircon age peaks and valleys comprise one population of epsNd(t) data and juvenile input in both decreases with time from 70-80% prior to 1500 Ma to 50-60% after this time. There is no preference for juvenile input during age peaks compared to age valleys. Most of the juvenile rocks in collisional orogens date to pre-collisional (accretionary), ocean-basin closing stages and not to the collision between cratons. The most widely preserved tectonic setting for both juvenile and reworked crust in collisional orogens is the continental arc. The Nd isotopic data can be interpreted in two ways: 1) juvenile continental crust is produced before, during and after craton collisions, but only during craton collisions is it preserved in large volumes, or 2) zircon age peaks reflect periods of accelerated growth of continental crust. In both models, the zircon age peaks could represent periods of enhanced granitoid production due to faster subduction driven by accelerated mantle convection. In the first model the peaks are preservation peaks, whereas in the second model they are crustal growth peaks.