Roberta Rudnick on the Continental Crustal Composition Paradox

Listen to the podcast here or wherever you get your podcasts.


The average composition of continental crust approximates that of an intermediate rock such as andesite. But mantle-derived magmas are basaltic in composition, with less silica and more magnesium and iron. Roberta Rudnick was one of the first to recognize this discrepancy. In the podcast, she describes the measurements that led us to identify what has come to be known as the continental crustal composition paradox. She then explains the various theories that purport to resolve the paradox. While some are more speculative than others, she thinks all of them probably play a role in resolving the paradox.

Rudnick is a Distinguished Professor in the Department of Earth Science at the University of California Santa Barbara.


Podcast Illustrations

Images courtesy of Roberta Rudnick unless otherwise noted.


Continental Crust

Continental crust is conventionally divided into three layers — upper, middle, and lower. The layers are defined by the variation of seismic wave speed (Vp) with depth, as indicated in the figure. The divisions are somewhat arbitrary, as the seismic wave speed varies continuously, albeit non-linearly, with depth.

Rudnick, R. L. & Fountain, D.M. (1995), Reviews of Geophysics, 33, 3

Measuring the Composition of the Upper Crust

As Rudnick describes in the podcast, one way of determining the average composition of the upper crust is simply to measure the composition at many locations. In one study, 14,000 grid samples were analyzed from the Canadian Shield for major and a few trace elements. Such studies show that major elements may vary by a factor of two, but trace elements vary by orders of magnitude.

Space shuttle view of Thunder Bay, Ontario
Grid measurements by Eade et al. (1973), grid for illustrative purposes only


Analyzing Shales

In the podcast, Rudnick says that fine-grained sedimentary rocks, such as the shales pictured here (at top), have trace-element compositions that are reflective of the upper crust as a whole, but only for the insoluble elements. The soluble elements, such as magnesium, calcium, and sodium, have been weathered out and transported to the oceans.

Mancos Shale, Utah; photo: J. St. John


Studying the Lower Continental Crust

Granulite Terranes

Granulites are highly metamorphosed rocks that come from the lower crust. They are brought to the surface through uplift, erosion, and tectonic transport along faults. The image shows granulites that, before metamorphosis, were fine-grained, clay-rich sedimentary rocks.

Mauricio Mazzuchelli, Ivrea Zone, Italy


Xenoliths

Xenoliths are rocks that have been brought up from the lower crust or the mantle as fragments in explosively erupting volcanoes or in the deep volcanic pipes called kimberlites, which also contain diamonds.

Professor Shukrani Manya, University of Dar es Salaam, Tanzania

Bill  McDonough, Queensland, Australia

Professors Gao and Wu, Shanxi, China


Seismic Wave Velocities

As she explains in the podcast, seismic wave velocities provide a way of studying the lower continental crust remotely. In this plot, the velocity of seismic p-waves is plotted against the silica content of rocks that characterize the upper crust (blue), the middle crust (red), and the lower crust (green). The speeds increase with depth by an amount that cannot be explained just by the effects of metamorphosing surface rocks but requires a change in the major element composition of the rocks.

Huang, Y. et al. (2013), Geochemistry, Geophysics, Geosystems 14(6): 2003, doi: 10.1002/ggge.20129


Composition of Continental Crust

Quite remarkably, the measurements of continental crust composition made by F.W.Clarke over 100 years ago agree quite well with modern estimates.

Rudnick et al. (2003), pub. Elsevier Ltd.,Treatise on Geochemistry, vol. 3, pp. 1-64

Clarke, F. W. (1889), Phil. Soc. Washington Bull. Vol. XI pp. 131-142


Trace-Element Signatures

Trace-Element Composition of Continental Crust

The plot shows the average abundances of rare earth elements and other selected trace elements in average continental crust. The abundances are normalized to the corresponding abundances in the primitive mantle. The data show a relative depletion of Nb (niobium) and a relative enrichment of Pb (lead). This signature matches that of lavas generated by the magmatism above subduction zones that creates island arc volcanoes.

Trace-Element Composition of Basalts

Trace-element abundances in basalts normalized to the primitive mantle abundances. The abundances in island arc basalts (IAB) show a relative depletion of Nb and a relative enrichment of Pb, which is similar to what we see in the continental crust, as shown in the plot above. By contrast, basalts produced at ocean islands above mantle plumes, such as Hawaii (OIB), and at mid-ocean ridges (MORB) do not show this signature. As Rudnick says in the podcast, this strongly suggests that continental crust is formed in island arc volcanoes that erupt above subduction zones.

Courtesy of the Open University


Island Arc Volcanoes

Illustration of a volcanic arc and its relationship to a subduction zone.

Courtesy of the Open University

Example of island arc volcanoes above a subduction zone: the Aleutian islands above the Aleutian subduction zone.

Photo: John Lyons/USGS


Explaining the Continental Composition Paradox

I: Lower Crustal Removal — Density Foundering

One way to convert material with a basaltic composition into an andesitic (intermediate) composition, which is what we measure for continental crust taken as a whole, is to remove a mafic component. In the podcast, Rudnick describes how a dense continental rock could break off and sink into the mantle. This idea was also discussed by Peter Molnar in his episode as a possible contributor to the extreme elevation of present-day Tibet. continental crust = pale green; orange = lithospheric mantle; blue = eclogite; dark green = asthenospheric mantle.

I: Lower Crustal Removal — ‘Relamination’

Relamination posits that when continental plates subduct, the downgoing plate partially melts and separates into more felsic material that is buoyant and more mafic material that is denser. The former rises and is incorporated into the base of the continental crust, thereby making the crust more silica-rich overall. The denser material sinks into the mantle. In the figure, the downgoing plate should be labeled as a continental plate rather than as an oceanic plate. Relamination has been suggested for other tectonic settings as well, such as when sediment accumulated on an oceanic plate is subducted, when an island arc is subducted, or when a fore-arc or accretionary prism is subducted.

Hacker, B.R. et al. (2015), Annual Review of Earth and Planetary Science 43:167–205

II: Hot subduction

One of the theories discussed in the podcast invokes the supposition that during the Archean, the lithosphere was hotter. Hot basaltic crust melts before it dehydrates, and the first material to melt, which has a granitic composition, rises and becomes part of the crust of an overriding continent, either as lava on the surface or as granite at depth.

Animation courtesy of T. Atwater, UCSB

III: Chemical Weathering

The amount of Mg in continental crust can be reduced by chemical weathering. Water dissolves the soluble elements, which include Mg, Ca, and Na. These wash into the sea, where they are absorbed by oceanic crust, which eventually subducts. During subduction, the Mg is retained within the subducting slab and transported into the mantle, while Na and possibly Ca return to the crust via magmatism at volcanic island arcs.