1081st General Monthly Meeting and the 50th Clarke Memorial Lecture
Origins of the Continental Crust
Professor Richard J. ARCULUSDepartment of Geology, Australian National University
Date: June 5th, 1999
ABSTRACT
The genesis of the continental crust is a first-order geological and geochemical problem. We have some understanding of the bulk composition of the crust after decades of effort and a variety of chemical and physical approaches. To a first approximation, the marked enrichment of a number of alkali elements (~100 fold) in the continental crust compared with estimates of primitive mantle abundances, is matched by complementary depletions in the global mantle source regions tapped by mid-ocean ridge basalts. In other words, the "depleted mantle" source of the most abundant crustal rock type (oceanic) represents a residue from continental crust extraction. In fact, depending on the primitive mantle abundances chosen, it is probable that the whole of the mantle has been involved in formation of the continental crust. Consequently, there is a strong geochemical argument that any depleted ("refractory") mantle component of the continental lithosphere can only represent a small fraction of the total mantle involved in continental crust formation. There can be no enduring retention of residual mantle in any vertical, continental crust-mantle differentiation process, and any juxtaposition may likely represent a fortuitous linkage of buoyant crust and mantle.
We know that the granodioritic-to-tonalitic major element composition of the continental crust can be produced in a variety of ways: (1) fractional crystallisation of basalt; (2) direct melting of subducted/underplated mafic lithosphere; (3) fractionation of high-Mg andesites generated by wet melting of peridotite. And in detail, it is clear that much of the crust, while produced initially in the Archean and early Proterozoic, has been subjected to considerable intra-crustal recycling.
Trace element and isotopic systematics require (at some stage) in continental crust genesis: (1) low % of mantle partial melting; (2) garnet-bearing protolith involvement; (3) fluid/solid elemental fractionation to account for the marked enrichment of Pb compared with other trace elements of similar melt/solid incompatibility.
Temporal changes in continental crust genesis through Earth history have probably been important. Unquestionable continental crustal growth has taken place in the Phanerozoic (e.g., much of eastern Australia east of the Rodinia break-up margin). Post-Proterozoic constraints are:
1) without deep recycling of H2O, neither melting of a subducted slab nor overlying mantle wedge occurs. Furthermore, the order of crystalline phase appearance in cooling basalt magmas containing a few % H2O is: olivine, pyroxene, plagioclase-the delay in plagioclase appearance in the "wet basalt case" enhances concentration of Si, Al, Na, and K in residual magmas (key elements present in abundance in granites). The presence of hydrated minerals (amphibole/biotite) in crystallisation products of wet magmas is also important in generation of granitic magmas during ultra-metamorphism of continental crust.
2) a key requirement for fractional crystallisation is for the system to "be left alone" (ie allowed to cool and crystallise) with minimal accompanying magma chamber tapping or replenishment. Neglect is a function of local magma flux, and in arcs, this is ~0.1 of the ridge flux.
3) disposal of ultramafic/mafic crystalline cumulates from basalt fractionation
processes is crucial for production of granodioritic crust. Cumulates may be
underplated (sub-Moho) as part of the continental lithosphere or dynamically
removed. Melts escaping from a supra-subduction zone system are
In detail however, it is also clear that the trace element compositions of modern arc magmas differ significantly from those of the continental crust, the former being markedly enriched in alkali and alkaline earth elements (plus Pb) compared with the rare earths. It is likely that a combination of several potential protoliths is required including: (1) wet partial melts of high-grade (eclogite, garnet amphibolite) mafic lithologies; (2) wet partial melts of the mantle involving recycled, subducted slab-derived H2O; (3) some minor input of intraplate (plume-derived) magmas.
The Toast is: The Continental Crust
A report on the 50th Clarke Memorial Lectureby Dr Edmund Potter
The Society's 50th Clarke Memorial Lecture was delivered at Macquarie University on 2nd June 1999 by renowned geophysicist Professor Richard J.Arculus (Australian National University) on the subject, "Origins of the Continental Crust". The Society's President, Assoc.Prof. Tony Baker, was in the Chair before an audience of members and visitors numbering nearly fifty.
Aided by a continuous display of uncluttered and colourful illustrations, Professor Arculus confirmed his enviable grasp of crustal facts and interpretation in a 60-minute effusion of verbal volcanism.
Only in the past decade of earth's 5 billion years has the option emerged of our living apart from the earth's crust. In this same decade eruptions, earthquakes, and tsunamis have reminded us that the crust on which we live is replete with weaknesses betraying the vast shell of restless incandescence not far beneath our feet.
And yet, as sightseers to the Grand Canyon soon appreciate, earth's crust is not like some even-textured concrete raft but is a grab-bag of crushed allsorts. Furthermore, the water-immersed part of the crust differs in composition from the protruding Continental crust. For example, rubidium and caesium, elements notable for the water solubility of their compounds, are some fifty or more times less abundant in the Oceanic crust. The obvious explanation is, however, deceptive, because salt water continually leaks deeply beneath the oceanic crust, so that over the aeons these two elements have been becoming fixed indissolubly in mineral aluminosilicates.
Professor Arculus dwelt extensively on the formation of granite, the major component of the Continental crust. The granite (he argued) did not come from the mantle by a melting/solidification cycle. Its possible sources include fractional crystallization of basalt, direct melting of subducted mafic lithosphere, and fractionation of high-magnesium andesites. The jury still seems to be out, but certainly water is essential for granite formation, justifying a role for volcanoes which, in recycling the oceans act as the Continental crust factory. Estimates of the overall material fluxes support the conclusion that while no new crust may be forming it is being recycled.
At question time the pace slackened but continued to reveal how precarious are the paths to geological discovery, as illustrated by the following samples (A=Arculus).
Q. Evidently you prefer the steady state model of the Continental
crust?
A. Yes, one accepts the model until it suffers a crucial setback.
Q. Do we know enough about the role of water?
A. No, we don't know the water balance for instance.
Q. How much do meteorites contribute over time?
A. Don't know, but our moon's early and still-visible encounters may
be a guide.
(Our thanks go to Edmund Potter for the preparation of this report.)