Disparities in oxygen isotopes of detrital and igneous zircon identify erosional bias in crustal rock record

Abstract

The detrital zircon oxygen isotope record has been used to describe 4.4 billion years of crustal evolution and maturation. Implicit in the application and use of large databases of oxygen isotope ratio (δ18O) from detrital zircon is that such data are more representative of the history of the continental crust than are exposed rocks. Here we show that the δ18O of igneous zircon in preserved rocks varies greatly from the detrital zircon record over time, both in terms of the magnitude of δ18O and the secular variation. Starting from 4.0 Ga and lasting throughout the Archean, the detrital zircon record shows consistently higher δ18O values than the rock record. In post-Archean times, detrital zircon δ18O is conspicuously higher during periods associated with supercontinent amalgamation. Differences between detrital and igneous zircon δ18O records imply that the igneous source rocks from which high δ18O detrital zircon are derived have been entirely eroded and are not represented in the known geological record. We postulate that the ‘missing’ record of high δ18O magmatic rocks consists of upper crustal sedimentary-derived melts from collisional belts, such as that found in the Himalayan orogen. The observation of higher δ18O in Archean detrital zircon further requires a revision of the current paradigm for Archean magmatism. More broadly, recognition of a ‘missing’ component of the upper crustal rock record calls into question the degree to which the detrital zircon record accurately reflects the composition of the bulk continental crust.