Geol 08v36n8 Archean subduc.pdf

Did the character of subduction change at the end of the Archean?

Constraints from convergent-margin granitoids

Kent C. Condie

Department of Earth and Environmental Science, New Mexico Institute of Mining and Technology, Socorro, New Mexico 87801, USA

ABSTRACT

Large ion lithophile and high ﬁ eld strength element distributions in juvenile upper

continental crust are controlled chieﬂ y by the abundance of tonalite-trondhjemite-

granodiorite (TTG) in the Archean shifting to a combination of TTG, calc-alkaline

granitoid, and graywacke control thereafter. Geochemical differences between TTG and

high-silica adakites do not require production of most TTG magmas in descending slabs.

Changes in the ratio of TTG to calc-alkaline granitoids after 2.5 Ga indicate that Archean

subduction zones must have differed from younger subduction zones in two very impor-

tant ways: (1) a deep maﬁ c crust served as a TTG magma source (either as thickened

crust or in descending slabs), and (2) they did not give rise to signiﬁ cant volumes of calc-

alkaline magma. Thickened maﬁ c crust in the Late Archean may have resulted from plate

jams in subduction zones caused by thicker oceanic crust and oceanic plateaus produced

during Late Archean mantle thermal events.

Keywords : continental crust, tonalite-trondhjemite-granodiorite, TTG, subduction, late Archean.

INTRODUCTION

Although chemical changes in the composition of juvenile con-

tinental crust were proposed in the 1970s, it was not until the 1980s

or later that most of these changes became widely recognized (Ronov,

1972; Engel et al., 1974; Taylor and McLennan, 1981, 1985; Condie,

1993). Compositional differences between Archean and later ﬁ ne-grained

detrital sediments are also well documented, although their interpreta-

tion is still controversial because of possible tectonic setting biases and

recycling effects (McLennan et al., 2006). Increases in large ion litho-

phile elements (LILE) and changes in related element ratios (such as

K 2 O/Na 2 O, La/Yb, and Sr/Y) in juvenile post-Archean upper continen-

tal crust are now reasonably well established, although their signiﬁ cance

in terms evolving tectonic regimes on Earth is a matter of discussion

(Taylor and McLennan, 1985; Gibbs et al., 1986; Condie, 1993; Rudnick

and Fountain, 1995; McLennan et al., 2006). Because the composition

of new upper continental crust is largely controlled by the composition of

convergent-margin granitoids (Condie, 1993), the relative volume

and origin of the tonalite-trondhjemite-granodiorite (TTG) suite are

critical to understanding changing crustal composition at the Archean-

Proterozoic boundary. Although the importance of TTG production dur-

ing the Archean is well known, it is now recognized that TTGs have been

produced at subduction zones from the Archean onward (Condie, 1993,

2005; Smithies, 2000; Kay et al., 2005). However, the question of whether

Archean TTGs were produced in a different tectonic setting (such as

melting of the roots of oceanic plateaus) from later TTGs is still debated

(Kamber et al., 2002; Martin et al., 2005; Condie, 2005; Nair and Chacko,

2005; Bedard, 2006; Foley, 2008).

Results of this study suggest that geochemical differences between

TTG and high-silica adakites do not require that all TTG magmas be pro-

duced in descending slabs, as suggested by Martin (1999). Increases in

LILE and high ﬁ eld strength elements (HFSE) in post-Archean juvenile

upper continental crust reﬂ ect chieﬂ y an increase in the ratio of calc-

alkaline to TTG magma production. This, in turn, implies that Archean

subduction zones differed from later subduction zones by producing thick

maﬁ c lower crust as a source for many TTG magmas, and by not produc-

ing signiﬁ cant volumes of calc-alkaline magma.

TABLE 1. ESTIMATES OF TONALITE-TRONDHJEMITE-GRANODIORITE/

CALC-ALKALINE RATIO WITH TIME

>2800

Age

(Ma)

2800–2500 2500–1000

<1000

Map TTG/CA

8 (4–10)

0.10

0.37

By location

TTG

TTG/CA

5.6

12.5

0.92

0.70

By number of samples

TTG

386

328

138

171

169

242

374

TTG/CA

2.7

9.4

0.54

0.65

TTG/CAassumed

0.5

0.6

Note : TTG—tonalite-trondhjemite-granodiorite suite; CA— calc-alkaline.

NEW AVERAGES

Table 1 shows three estimates of the ratio of TTG to calc-alkaline

granitoids (TTG/CA ratio) in four time windows. The map-scaling

estimate was described in Condie (1993), where the calc-alkaline suite

was not distinguished from the TTG suite. TTG/CA ratios in Table 1

are also given by number of locations in which each is reported and by

the total number of samples chemically analyzed. Although the three

estimates for a given age window can vary by as much as a factor

of 10, most are within a factor of 2. TTG/CA ratios used in weight-

ing each rock type for the new average upper crustal compositions are

given in Table 1.

Average chemical compositions of TTG and calc-alkaline plutonic

suites are given in GSA Data Repository Tables DR1 and DR2 1 , together

with median values and one standard deviation of the mean. The distinc-

tion between calc-alkaline and TTG suites is based on major element

trends such as those on the K-Na-Ca graph (Martin, 1994), various silica

variation diagrams, Sr content, Eu anomaly (Eu/Eu*) distribution, and low

1 GSA Data Repository item 2008148, Tables DR1–DR4 and Figures

DR1–DR4, geochemical data and additional ﬁ gures, is available online at www.

geosociety.org/pubs/ft2008.htm, or on request from editing@geosociety.org or

Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.

Geology , August 2008; v. 36; no. 8; p. 611–614; doi: 10.1130/G24793A.1; 2 ﬁ gures; 1 table; Data Repository item 2008148.

611

GEOLOGY, August 2008

TTG

Calc-Alkaline

Average UC

contents of heavy rare earth elements and Y. Because of restite plagioclase

or plagioclase fractional crystallization, calc-alkaline granitoids have

distinctly low Sr contents and signiﬁ cant negative Eu anomalies, very dis-

tinct from TTGs (Martin, 1994; Martin et al., 2005). Sanukitoids are not

included in the database; they were eliminated using the geochemical ﬁ lter

suggested by Martin et al. (2005).

Revised estimates of upper continental crustal composition (from

Condie, 1993) are given in Tables DR3 and DR4. Also given are rock

weighting factors and average compositions of each rock type. Other

than the new TTG and calc-alkaline categories, most of the weight-

ing factors and average rock compositions are similar to those given

in Condie (1993). Methods and error estimates were also discussed in

Condie (1993).

1.5

0.5

500 1000 1500 2000 2500 3000 3500 4000

RESULTS

Although there are minor within group secular changes in the aver-

age composition of TTGs and calc-alkaline granitoids (Tables DR1 and

DR2), almost all of these are within one standard deviation of the grand

mean values. Such changes do not contribute signiﬁ cantly to the changes

in the composition of upper continental crust at the end of the Archean. As

shown in Tables DR3 and DR4, LILE and HFSE (Nb, Ta, Zr, Hf) distribu-

tions in the upper continental crust are controlled chieﬂ y by TTG in the

Archean, shifting to a combination of TTG, calc-alkaline granitoid, and

graywacke thereafter. Except for Sr and trace metals, average graywacke

is similar to average TTG for most elements. Signiﬁ cant increases in

LILE and HFSE and a decrease in Sr between the 2.8–2.5 Ga and the

2.5–1.0 Ga time windows reﬂ ect, in large part, a shift from TTG control

to calc-alkaline control. This is also evident in key element ratios such

as K 2 O/Na 2 O, (La/Yb) n , Sr/Y, Th/U, Eu/Eu*, and Nb/Ta (Fig. 1; Tables

DR3 and DR4; Figs. DR1–DR4). An increase in K 2 O/Na 2 O ratio in upper

crust at the end of the Archean is clearly controlled by the distribution of

average compositions of TTG and calc-alkaline granitoids in Figure 1A

(K 2 O/Na 2 O > 0.5 for calc-alkaline suite and <1 for most TTGs). Likewise,

decreases in (La/Yb) n and Sr/Y ratios in early Proterozoic upper crust

reﬂ ect an increasing calc-alkaline component (Figs. 1B, 1C). Although

both La and Yb contents control the La/Yb ratio, Yb increases faster than

La after 2.5 Ga (due to a drop in garnet content of TTG sources), thus

the La/Yb ratio decreases. Similarly, the large drop in Sr and Eu/Eu* in

the calc-alkaline component (due to restite plagioclase and/or fractional

crystallization) controls the decreases in upper crustal Sr/Y and Eu/Eu*

ratios with time (Fig. 1C; Tables DR1 and DR2; Fig. DR1). Th and U

both increase in the upper crust at the end of the Archean, but the Th/U

ratio remains constant or drops slightly (Fig. DR2). The Nb/Ta ratio in

upper continental crust remains roughly constant from the Early Archean

onward (Tables DR3 and DR4; Fig. DR3).

Age (Ma)

100

TTG

Calc-Alkaline

Average UC

500 1000 1500 2000 2500 3000 3500 4000

Age (Ma)

TTG

Calc-Alkaline

Average UC

100

500 1000 1500 2000 2500 3000 3500 4000

DISCUSSION

A major unresolved issue regarding compositional changes near the

end of the Archean is that of whether Archean TTGs are really high-silica

adakites, and thus require a magma source in descending slabs (Kamber

et al., 2002; Martin et al., 2005; Condie, 2005). Although Martin and

colleagues made a case for a TTG-adakite connection based on various

element distributions (Martin, 1999; Martin et al., 2005), key element

indicators such as Mg#, Ni, Cr, Nb/Ta, Pb/Nd, and B/Be do not favor

such a connection for many (or most) TTGs (Smithies, 2000; Kamber

et al., 2002; Condie, 2005; Foley, 2008). Although the database of Martin

and Moyen (2002) (with no rocks younger than 2.5 Ga) suggests MgO

contents of TTGs similar to high-silica adakites, my extensive database

(Tables DR1 and DR2) clearly shows that most TTGs have Mg numbers

(<50) less than high-silica adakites (also summarized in Condie, 2005,

Fig. 5 therein). As pointed out by Kamber et al. (2002), the Nb/Ta ratio of

adakites is ≥16, but only a few of the 38 TTG averages have ratios above

Age (Ma)

Figure 1. A: Secular distribution of K 2 O/Na 2 O in average convergent-

margin granitoids and in average upper continental crust (UC). Data

are averages from Tables DR1–DR4 (see footnote 1). Vertical bars are

uncertainty estimates expressed as one standard deviation of mean

values. B: Secular distribution of (La/Yb) n in average convergent-

margin granitoids and in average UC. Data are averages from Tables

DR1–DR4. Vertical bars are uncertainty estimates expressed as one

standard deviation of mean values; n—normalized to primitive mantle

(Sun and McDonough, 1989). C: Secular distribution of Sr/Y in aver-

age convergent-margin granitoids and in average upper continental

crust (UC). Data are averages from Tables DR1–DR4. Vertical bars are

uncertainty estimates expressed as one standard deviation of mean

values. TTG—tonalite-trondhjemite-granodiorite.

612

GEOLOGY, August 2008

16 (Tables DR1 and DR2; Figs. DR3 and DR4). Considering these key

element indicators, it appears that only a relatively small number of TTGs

of any age have similarities to high-silica adakites.

The results of this and previous studies clearly agree that Late

Archean TTGs must be derived from wet maﬁ c sources in which

amphibole ± garnet remains in the restite (Martin, 1994; Rapp, 1997;

Moyen and Stevens, 2006; Bedard, 2006). This means that the maﬁ c

source must be deeper than post-Archean maﬁ c sources, and the cur-

rent debate centers around what caused this deeper Archean source.

Martin (1986, 1994, 1999) has long championed descending slabs as

this source by analogy with adakites. However, as pointed out above,

key geochemical indicators do not favor an adakite connection for many

TTGs, regardless of age. An equally viable possibility is that the thick-

ened maﬁ c source is produced by “plate jams” in Archean subduction

zones, during which maﬁ c crust is stacked up and thickened and the root

zones undergo partial melting to produce TTG magmas (Davies, 1992;

van Hunen et al., 2000; Nair and Chacko, 2005; Clemens et al., 2006).

Because this occurred during one or two Late Archean global mantle

thermal events (mantle plume events?), the upper mantle is hotter and

oceanic plates are more buoyant and tend to resist deep subduction. Parts

of oceanic plateaus may also be added to the accretionary stacks at con-

vergent margins, thickening the maﬁ c crust even more. In fact, Bedard

(2006) proposed a model whereby the thickened root zones of oceanic

plateaus may give rise to TTG magmas without subduction. Locally

preserved eclogite–high-pressure granulite metamorphic assemblages in

Late Archean greenstones (Brown, 2007) support the existence of thick

maﬁ c crust (>50 km) in Late Archean arc systems.

Although not part of this study, sanukitoids (and Closepet-type

granites) become increasingly abundant in the latest Archean (2.6–2.5 Ga).

As pointed out by Martin et al. (2005), this may result from cooling of the

mantle such that slab-derived felsic melts were completely consumed by

reactions in the mantle wedge, followed by partial melting of the wedge

to produce sanukitoid magmas. This suggests that the change from TTG

to calc-alkaline magmatism after 2.5 Ga may have involved a transition

period during which sanukitoid magma production was widespread.

The next question is why the depth of maﬁ c sources at convergent

margins decreased after the Archean. The model of Martin (1994, 1999) is

still most effective, i.e., decreasing thermal gradients in subduction zones

leads to steeper subduction and thus to a greater volume of mantle wedge

(Peacock, 2003). Descending plates devolatilize and melting shifts to the

mantle wedge where calc-alkaline magmas are produced. Felsic calc-

alkaline components may form by either fractional crystallization or par-

tial melting of maﬁ c components underplated in the lower crust at depths

<50 km. Consistent with this idea is the fact that post-Archean continental

crust typically has a seismically high velocity layer in the lower crust,

which could represent underplated maﬁ c crust (Durrheim and Mooney,

1991). A major unresolved question with this model is how to cool sub-

duction zones so rapidly after 2.5 Ga.

Another important constraint is that mantle wedges must be

enriched in LILE before they can yield calc-alkaline magmas. Experi-

mental studies clearly indicate that calc-alkaline basaltic magmas can-

not be produced from depleted mantle sources (Sisson et al., 2005).

Two factors may contribute to enriching post-Archean mantle wedges

in LILE. Devolatilization of downgoing slabs alone may be sufﬁ cient

to carry LILE from descending slabs into the mantle wedge. However,

subduction of sediments derived from newly formed cratons in the Late

Archean also may contribute to the LILE inventory that eventually ends

up in mantle wedges.

Using the constraints discussed in this study, a diagrammatic sum-

mary of a possible tectonic-thermal history of Earth between 3 and 2 Ga is

given in Figure 2. The model assumes that subduction-related granitoids

during the Late Archean were derived from oceanic crustal sources (includ-

LILE enrichment

in continental crust

Depth of granitoid

sources

Continental crust

production rate

Subduction

distribution

Late Archean

mantle thermal

events

2000

2500

2700

3000

Age (Ma)

Figure 2. Schematic representation of possible tectonic-thermal his-

tory on Earth between 3 and 2 Ga (based on data in Davies, 1995;

Condie, 1998; Condie and Benn, 2006; Condie and Kroner, 2008).

LILE—large ion lithophile elements.

ing oceanic plateaus), whereas most post-Archean examples come from

maﬁ c underplates, which, in turn, come from mantle wedges. Although

plate tectonics began at least locally by 3 Ga and became widespread in

the Late Archean (Abbott et al., 1994; Condie and Benn, 2006; Condie

and Kroner, 2008), modern-style subduction zones did not become wide-

spread until ca. 2.2 Ga. Two major pulses of thermal activity in the upper

mantle ca. 2.7 and 2.5 Ga enhanced subduction rates and rates of produc-

tion of both continental and oceanic crust. Because of extensive plate jams

and accretion of oceanic plateaus in Late Archean subduction zones, maﬁ c

crust was greatly thickened and underwent partial melting to produce

TTG magma suites. Although LILE enrichment began at this time due to

delamination and devolatilization of maﬁ c components, widespread LILE

enrichment did not begin until the Early Proterozoic, when modern style

mantle wedges came into existence.

CONCLUSIONS

1. Compared to high-silica adakites, many Late Archean TTGs are

not enriched in Mg, Cr, or Ni, and have low Nb/Ta ratios, suggesting that

they are not derived from descending slabs.

2. Increases in LILE and HFSE, a decrease in Sr, and the appearance

of large negative Eu anomalies in juvenile upper continental crust after

the Archean reﬂ ect chieﬂ y a decrease in TTG magma production and a

corresponding increase in calc-alkaline magma production at convergent

plate margins.

3. To accommodate these geochemical changes, Late Archean sub-

duction zones must have differed from younger subduction zones in two

very important ways: (1) a deep maﬁ c crust served as a TTG magma

source (either as thickened crust or in descending slabs), and (2) they did

not give rise to signiﬁ cant volumes of calc-alkaline magma.

4. One way to thicken maﬁ c crust in the Late Archean is by plate jams

in subduction zones caused by thicker oceanic crust and oceanic plateaus

produced during Late Archean global thermal events. Modern subduction

zones became widespread in the Early Proterozoic as magma production

shifted to mantle wedges, where the calc-alkaline suite is produced. LILE

enrichment in mantle wedges reﬂ ects some combination of devolatiliza-

tion of descending slabs and sediment subduction.

GEOLOGY, August 2008

613

ACKNOWLEDGMENTS

I appreciate in-depth reviews by Herve Martin, Bor-ming Jahn, and Ali Polat.

Martin, H., 1994, The Archean grey gneisses and the genesis of the continental

crust, in Condie, K.C., ed., Archean crustal evolution: Amsterdam, Elsevier,

p. 205–259.

Martin, H., 1999, Adakitic magmas: Modern analogues of Archean granitoids:

Lithos, v. 46, p. 411–429, doi: 10.1016/S0024–4937(98)00076–0.

Martin, H., and Moyen, J.-F., 2002, Secular changes in tonalite-trondhjemite-

granodiorite composition as marker of the progressive cooling of Earth:

Geology, v. 30, p. 319–322.

Martin, H., Smithies, R.H., Rapp, R., Moyen, J.-F., and Champion, D., 2005, An

overview of adakite, tonalite-trondhjemite-granodiorite (TTG), and sanuki-

toid: Relationships and some implications for crustal evolution: Lithos,

v. 79, p. 1–24, doi: 10.1016/j.lithos.2004.04.048.

McLennan, S.M., Taylor, S.R., and Hemming, S.R., 2006, Composition, differ-

entiation and evolution of continental crust: Constraints from sedimentary

rocks and heat ﬂ ow, in Brown, M., and Rushmer, T., eds., Evolution and dif-

ferentiation of the continental crust: Cambridge, UK, Cambridge University

Press, p. 92–134.

Moyen, J.-F., and Stevens, G., 2006, Experimental constraints of TTG petro-

genesis: Implications for Archean geodynamics, in Benn, K., et al., eds.,

Archean geodynamics and environments: American Geophysical Union

Geophysical Monograph 164, p. 149–175.

Nair, R.K., and Chacko, T., 2005, Oceanic plateaus: Nuclei for Archean cratons:

Geological Society of America Abstracts with Programs, v. 37, no. 7, p. 494.

Peacock, S.M., 2003, Thermal structure and metamorphic evolution of subduct-

ing slabs, in Eiler, J., ed., Inside the subduction factory: American Geo-

physical Union Geophysical Monograph 138, p. 7–22.

Rapp, R.P., 1997, Heterogeneous source regions for Archean granitoids: Experi-

mental and geochemical evidence, in de Wit, M.J., and Ashwal, L.D., eds.,

Greenstone belts: Oxford, Clarendon Press, p. 267–279.

Ronov, A.B., 1972, Evolution of rock composition and geochemical processes in

the sedimentary shell of the Earth: Sedimentology, v. 19, p. 157–172, doi:

10.1111/j.1365–3091.1972.tb00019.x.

Rudnick, R.L., and Fountain, D.M., 1995, Nature and composition of the con-

tinental crust: A lower crustal perspective: Reviews of Geophysics, v. 33,

p. 267–309, doi: 10.1029/95RG01302.

Sisson, T.W., Ratajeski, K., Hankins, W.B., and Glazner, A.F., 2005, Voluminous

granitic magmas from common basaltic sources: Contributions to Mineral-

ogy and Petrology, v. 148, p. 635–661, doi: 10.1007/s00410–004–0632–9.

Smithies, R.H., 2000, The Archean tonalite-trondhjemite-granodiorite (TTG)

series is not an analogue of Cenozoic adakite: Earth and Planetary Science

Letters, v. 182, p. 115–125, doi: 10.1016/S0012–821X(00)00236–3.

Sun, S.S., and McDonough, W.F., 1989, Chemical and isotopic systematics of

oceanic basalts: Implications for mantle composition and processes, in

Saunders, A.S., and Norry, M.J., eds., Magmatism in ocean basins: Geo-

logical Society [London] Special Publication 42, p. 313–345.

Taylor, S.R., and McLennan, S.M., 1981, The composition and evolution of the

continental crust: Rare earth element evidence from sedimentary rocks:

Royal Society of London Philosophical Transactions, ser. A, v. 301,

p. 381–399.

Taylor, S.R., and McLennan, S.M., 1985, The continental crust: Its composition

and evolution: Oxford, Blackwell, 312 p.

van Hunen, J., van den Berg, A., and Vlaar, N.J., 2000, A thermal-mechanical

model of horizontal subduction below an overriding plate: Earth and

Planetary Science Letters, v. 182, p. 157–169, doi: 10.1016/S0012–821X

(00)00240–5.

REFERENCES CITED

Abbott, D., Drury, R., and Smith, W.H.F., 1994, Flat to steep transition in subduc-

tion style: Geology, v. 22, p. 937–940, doi: 10.1130/0091–7613(1994)022

<0937:FTSTIS>2.3.CO;2.

Bedard, J.H., 2006, A catalytic delamination-driven model for coupled genesis of

Archean crust and sub-continental lithospheric mantle: Geochimica et Cos-

mochimica Acta, v. 70, p. 1188–1214, doi: 10.1016/j.gca.2005.11.008.

Brown, M., 2007, Metamorphic conditions in orogenic belts: A record of secular

change: International Geology Review, v. 49, p. 193–234.

Clemens, J.D., Yearron, L.M., and Stevens, G., 2006, Barberton TTG magmas:

Geochemical and experimental constraints on source-rock petrology, pres-

sure of formation and tectonic setting: Precambrian Research, v. 151,

p. 53–78, doi: 10.1016/j.precamres.2006.08.001.

Condie, K.C., 1993, Chemical composition and evolution of the upper continen-

tal crust: Contrasting results from surface samples and shales: Chemical

Geology, v. 104, p. 1–37, doi: 10.1016/0009–2541(93)90140-E.

Condie, K.C., 1998, Episodic continental growth and supercontinents: A man-

tle avalanche connection?: Earth and Planetary Science Letters, v. 163,

p. 97–108.

Condie, K.C., 2005, TTGs and adakites: Are they both slab melts?: Lithos, v. 80,

p. 33–44, doi: 10.1016/j.lithos.2003.11.001.

Condie, K.C., and Benn, K., 2006, Archean geodynamics: Similar to or different

from modern geodynamics?, in Benn, K., et al., eds., Archean geodynamics

and environments: American Geophysical Union Geophysical Monograph

164, p. 47–59.

Condie, K.C., and Kroner, A., 2008, When did plate tectonics begin? Evidence

from the geologic record, in Condie, K.C., and Pease, V., eds., When did

plate tectonics begin on planet Earth?: Geological Society of America Spe-

cial Paper 440 (in press).

Davies, G.F., 1992, On the emergence of plate tectonics: Geology, v. 20,

p. 963–966, doi: 10.1130/0091–7613(1992)020<0963:OTEOPT>2.3.CO;2.

Davies, G.F., 1995, Punctuated tectonic evolution of the earth: Earth and Plan-

etary Science Letters, v. 136, p. 363–379,

Durrheim, R.J., and Mooney, W.D., 1991, Archean and Proterozoic crustal evolu-

tion: Evidence from crustal seismology: Geology, v. 19, p. 606–609, doi:

10.1130/0091–7613(1991)019<0606:AAPCEE>2.3.CO;2.

Engel, A.E.J., Itson, S.P., Engel, C.G., Stickney, D.M., and Cray, E.J., 1974,

Crustal evolution and global tectonics: A petrogenetic view: Geological

Society of America Bulletin, v. 85, p. 843–858, doi: 10.1130/0016–7606

(1974)85<843:CEAGTA>2.0.CO;2.

Foley, S., 2008, A trace element perspective on Archean crust formation and on

the presence or absence of Archean subduction, in Condie, K.C., and Pease,

V., eds., When did plate tectonics begin on planet Earth?: Geological Soci-

ety of America Special Paper 440 (in press).

Gibbs, A.K., Montgomery, C.W., O’Day, P.A., and Erslev, E.A., 1986, The Archean-

Proterozoic transition: Evidence from the geochemistry of metasedimen-

tary rocks of Guyana and Montana: Geochimica et Cosmochimica Acta,

v. 50, p. 2125–2141, doi: 10.1016/0016–7037(86)90067–0.

Kamber, B.S., Ewart, A., Collerson, K.D., Bruce, M.C., and McDonald, G.D.,

2002, Fluid-mobile trace element constraints on the role of slab melting and

implications for Archean crustal growth models: Contributions to Mineral-

ogy and Petrology, v. 144, p. 38–56.

Kay, S.M., Godoy, E., and Kurtz, A., 2005, Episodic arc migration, crustal thick-

ening, subduction erosion and magmatism in the south-central Andes: Geo-

logical Society of America Bulletin, v. 117, p. 67–88.

Martin, H., 1986, Effect of steeper Archean geothermal gradient on geochemis-

try of subduction-zone magmas: Geology, v. 14, p. 753–756, doi: 10.1130/

0091–7613(1986)14<753:EOSAGG>2.0.CO;2.

Manuscript received 18 January 2008

Revised manuscript received 9 April 2008

Manuscript accepted 11 April 2008

Printed in USA

614

GEOLOGY, August 2008

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: