A REVIEW OF U-PB ZIRCON AGES FROM THE ROANOKE RAPIDS TERRANE, EASTERN PIEDMONT PROVINCE, VIRGINIA, AND NEW ND ISOTOPIC DATA BEARING ON THE NATURE OF CRATONIC BASEMENT

The Carolinia domain represents the southernmost tract of peri-Gondwanan crust in the Appalachian orogen. It is comprised of at least a dozen individual terranes, but many have only been characterized (e.g., ages, isotopic compositions) at the reconnaissance scale. The Roanoke Rapids terrane (RRT) in Virginia and North Carolina is the easternmost of these terranes. We provide here an update on U-Pb zircon ages (CA-ID-TIMS) from the Virginia portion of the terrane (some of which are revised from ages previously reported by us). The oldest protolith age (672.6 ± 1.1 Ma) is recorded in a screen or roof pendant of felsic metavolcanic rock hosted within a tonalitic pluton dated at 613.6 ± 0.9 Ma. Two other granitoid plutons yielded ages of 628.6 ± 3.6 Ma and 625.0 ± 0.6 Ma, respectively, whereas a metagabbro pluton is dated at 609.5 ± 0.8 Ma. The youngest age, from another felsic metavolcanic rock, is 551.5 ± 0.6 Ma. The intervals of igneous activity ~670 and ~630-610 Ma are similar to the few previously reported ages by others in the RRT. The RRT thus preserves a record of at least three magmatic pulses, and is the only Carolinia terrane with crust >670 Ma. New Nd isotopic results yield mostly negative initial ε_Nd values (-4.3 to -1.7), although the youngest rock has a positive (but still low) value of 1.8. Depleted mantle model ages range from ~1.1 to 1.6 Ga. Collectively, these results suggest the significant involvement of evolved crustal material in the petrogenesis of these rocks, i.e., the RRT was built on a basement of old (Mesoproterozoic) crust. Because the age range of RRT plutonic rocks matches well with the timing of magmatism in the ~630-610 time interval in the Virgilina sequence of the Carolina terrane, we (Owens and Hamilton, 2013) suggested that the two terranes may have been linked at that time. However, Virgilina sequence rocks yield uniformly positive initial ε_Nd values (Samson et al., 1995), consistent with primarily juvenile sources. Thus, the two terranes appear to have been constructed on fundamentally different substrates. The dominantly negative ε_Nd values for the RRT, even for the oldest rocks, suggest that it may have accreted to the margin of West Gondwana by this time. In contrast, the juvenile character of the Virgilina sequence implies that the Carolina terrane may still have been outboard of the Gondwanan margin.