A Matter of Preservation

See allHide authors and affiliations

Science  02 Jan 2009:
Vol. 323, Issue 5910, pp. 49-50
DOI: 10.1126/science.1168549

The fossil record is well known to be biased by the unevenness of the geographical and stratigraphical sampling effort and inequality in the rock record available for sampling (1). There is increasing evidence that a similar unevenness biases the geological record of the generation and evolution of the continental crust. In a recent study, Campbell and Allen (2) reemphasized the link between peaks in the distribution of the uranium- lead (U-Pb) crystallization ages of the mineral zircon (which reflect the ages of the parent igneous rocks) and the development of supercontinents. It is not clear, however, why the development of supercontinents should be associated with the generation of unusual volumes of igneous rocks. Instead, peaks of crystallization ages in the continental record are likely to reflect biases in preservation.

The bulk composition of continental crust is similar to that of rocks generated in subduction- related settings (3). Globally, subduction is continuous, which in turn suggests that the processes of crust generation should result in a continuum of ages. However, the preserved geological record is marked by peaks of crystallization ages that indicate peaks of magmatic activity (2) and by peaks of juvenile (mantle-derived) crust formation (4, 5) (see the first figure). If these peaks truly reflect the geological record, the implication is that continental growth is episodic on a global scale (6). However, the mechanism for this remains speculative (7). Alternatively, the peaks of both magmatic activity and juvenile crust formation may be a function of preservation, reflecting the unevenness of the rock record available for sampling. To what extent do magmas generated in different settings have different preservation potential? How does this preservation potential in turn shape the rock record now available for analysis?

Age is not everything.

The U-Pb age of the mineral zircon (calculated from its present-day U and Pb isotope ratios) corresponds to the age of crystallization of the parent igneous rock. Analysis of 7000 detrital zircons shows several peaks in their U-Pb crystallization ages over the course of Earth history (2) (bottom). Similar peaks are observed in the relative volumes of rocks of different ages that reflect juvenile crust (normalized to 100% for the total volume of crust) (4) (top). The peaks in both graphs often but not always coincide with the periods of supercontinent formation (gray bars in the bottom panel).

Individual magmatic events involve generation of new crust, reworking of preexisting crust, or a combination of the two processes. Thus, peaks of crystallization ages are not necessarily associated with periods of enhanced crustal growth (see the first figure). Nonetheless, several authors have linked peaks of magmatic activity to the development of supercontinents (2). As a result of plate tectonics, continental fragments periodically amalgamate to form large supercontinents. The igneous record associated with the development and breakup of supercontinents may be divided into three consecutive stages: subduction-related magmatism, collisional mountain building and magmas generated by crustal melting, and extensional magmatism (see the second figure). At issue is the extent to which the geological record of each stage will be preserved and represented in today's rock record.

Despite their similarity to the bulk composition of the continental crust (3), subduction-related rocks appear to have poor preservation potential. The global rates of removal of continental and island-arc crust through subduction into the mantle are similar to the crust generation rates at modern magmatic arcs (8). Instead, there is increasing evidence that subduction- related magmatic rocks are better preserved in extensional basins that lie inboard from the subduction zone (9).

Magmas associated with collisional mountain building are dominated by partial melting of preexisting crust. They are granitic in composition and, although the volumes generated may be small relative to other tectonic settings, it will tend to be protected within the enveloping supercontinent. It thus has good preservation potential in the geological record.

The extensional phase is dominated by basaltic magmatism (exemplified by the flood basalts and subordinate rhyolite associated with the breakup of Gondwana). The rocks may also be relatively sensitive to erosion into the oceans. Rocks from this phase do not contain large amounts of zircon and have relatively poor preservation potential.

These considerations show that preserved rocks tend to be from the end of the subduction- related and the collisional mountain building stages (see the second figure), because they are insulated in the cratonic interior and thus preserved from convergent plate margin erosion. Thus, the reason the continental evolution record is likely to be dominated by periods of supercontinent assembly is that these periods provide a setting for the selective preservation of crust. Age peaks reflect the interplay between zircon crystallization and preservation bias related to tectonic setting and do not require accelerated pulses of crust generation or magmatic activity. The peak between 2.9 and 2.5 billion years ago may be different, in that it coincides with the time of the stabilization of Archean crust as presently preserved and may thus preserve a more representative record of the rocks of those ages.

Preservation potential.

The volumes of magma generated (blue line), and their likely preservation potential (red lines), may vary in the three stages associated with the convergence, assembly, and breakup of a supercontinent. The preservation potential in the first stage is greater at margins where the subduction zone retreats oceanward to form extensional basins than at margins where the subduction zone advances toward the continent. Thus, peaks in the crystallization ages that are preserved (brown area) reflect the balance between the magma volumes generated in the three stages and their preservation potential.

Rocks formed before 2.9 billion years ago have an extremely poor preservation potential (see the first figure). Thus, the late Archean marks the transition from a period of uniformly poor preservation potential to one in which the geological record is biased by the tectonic setting in which the rocks were formed. The peaks of ages illustrated in the first figure have been interpreted in terms of episodic, rather than continuous, generation of continental crust (6, 7). Such models are largely based on the records of igneous and sedimentary rocks selectively preserved in “stable” areas and are therefore biased by the formation of supercontinents. The challenge now is to explore the geological records from stages with poor preservation potential.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
View Abstract

Stay Connected to Science

Navigate This Article