Brief Communications Arising | Published:

Asteroids and andesites

Nature volume 459, page E1 (28 May 2009) | Download Citation



Arising from: J. M. D. Day et al. Nature 457, 179–182 (2009)10.1038/nature07651; Day et al. reply

The production of terrestrial andesites in subduction zones is well established. Day et al.1 describe two examples of meteorites (GRA 06128 and GRA 06129) that they claim to represent “an entirely new mode of generation of andesite crust compositions” on asteroids; this suggestion has wide implications for the generation of andesitic planetary crusts in general. However, here we show that compositional data, particularly for the rare-earth elements (REEs) and other lithophile elements, presented in their paper1 do not substantiate this claim. We conclude that existing mechanisms for andesite generation do not need revision.


Day et al.1 present two relevant figures. The first (Fig. 1a in ref. 1) is a plot of Na2O plus K2O versus SiO2 that falls marginally within the andesite field. We consider that this figure could be misleading, as it is dominated by an exceptionally high Na2O content (6%) and should not have been used to infer that their samples have andesitic affinities.

The second figure (Fig. 1b in ref. 1) is a chondrite-normalized REE plot; we note that Day et al. also present data for the age, platinum group elements, Re–Os systematics and oxygen isotopes of the samples that do not directly address their thesis. REE values are very low (La varies between 0.95 and 1.48 p.p.m. or 2.6 to 4 times chondritic using volatile-free CI values for normalization). The chondrite-normalized REE patterns in these meteorites are flat but with a significant enrichment in Eu. These patterns are comparable, if a little lower than those of ‘Eu-rich’, albeit otherwise ‘normal’ eucrites2 (basaltic meteorites derived from the asteroid 4 Vesta) and do not at all resemble that of the ‘average terrestrial andesite’ REE pattern (or the majority of terrestrial arc andesites) that is also shown in the same figure. This pattern has a strong enrichment in the light REEs (for example, La = 16 p.p.m. (ref. 3) compared to 0.95–1.48 p.p.m. La for the meteorites). We believe that differences by factors of more than 10 need to be considered.

From the data given in the Supplementary Information of ref. 1, the sodium content (around 6%) is a factor of two higher, and the potassium content (about 0.2%) a factor of six lower, than averages for andesites or for the continental crust of the Earth. Indeed, the elevated silica in these rocks results from the high albite composition of the plagioclase, which dominates the mineralogy. In addition to the REEs, other significant lithophile elements such as Li, Rb, Sr, Cs, Zr, Hf, Nb, Ba, Pb, U and Th also show major discrepancies compared to their values in the continental crust, with the most incompatible of these (Th, Cs) being depleted by factors of 40 or more.

So although these meteorites are very interesting examples with high silica values (around 55%) that arose early (4.52 Gyr) on an asteroid, their compositions are fundamentally different from terrestrial andesites, and this study does not demonstrate that andesitic crusts resembling anything like the continental crust of the Earth might have evolved on such asteroids. Labelling these samples as ‘andesite’ carries the implication that the origin of these rocks has lessons for the formation of the continental crust of the Earth or elsewhere, and could be misleading.


  1. 1.

    et al. Early formation of evolved asteroidal crust. Nature 457, 179–182 (2009)

  2. 2.

    & Geochemistry of diverse lithologies in Antarctic eucrites. Lunar Planet. Sci. XIX, abstr. 790-1 (1988)

  3. 3.

    & Planetary Crusts: Their Composition, Origin and Evolution Table 11.4 (Cambridge Univ. Press, 2009)

Download references

Author information


  1. *Research School of Earth Sciences, Australian National University, Canberra, Australian Capital Territory 0200, Australia.

    • Richard Arculus
    • , Ian H. Campbell
    •  & Stuart Ross Taylor
  2. †Department of Geosciences, State University of New York at Stony Brook, Stony Brook, New York 11794-2100, USA

    • Scott M. McLennan


  1. Search for Richard Arculus in:

  2. Search for Ian H. Campbell in:

  3. Search for Scott M. McLennan in:

  4. Search for Stuart Ross Taylor in:

About this article

Publication history






By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing