Articles
A.Y.Glikson
Bureau of Mineral Resources
Geology and Geophysics
Canberra
ACT. Australia
Summary
The elucidation of the nature ofthe crust
which originally underlay and delimited
Archean volcanic-sedimentary green:
stone belts s essential for the under-
‘standing of the early Precambrian
‘crustal record -and through it of early
‘etrestria evolution, Stratigraphic evi-
dence from the Pilbara and Yilgarn
Coratons in Western Australia is consist
tent with datatrom the eastem Transvaal
Rhodesia and southern india, suggest
ingatundamental dichotomy of maior
greenstone sequences. Twoassemb-
lages are recognized, including: 1) early
greenstones which predate, and occut
as xenoliths within, the isotopically
‘oldest granites in each craton, and2)
late greenstones, which overlie the early
‘greenstones through major paracontor=
rmities accompanied by thick chert and
acid volcanic units, and may in places
overlap the early granites unconform-
ably. The early granites typically consist
‘of tonalite and granodioxite whose
‘geochemical andisotopic parameters
inaicate derivation trom parental basic
materials. In this article, the possible
relevance of these observations tothe
‘question of granite-groonsione relations
in the Canadian Precambrian Shield is.
considered, with particular attention to
the Archean crustal foundation of
Keewatin (Abitibi) and Yellowknite
greenstone bets.
Nature of the Problem
‘Two broad schools of thought exist
regarding the nature of the crust on
which Archean greenstones were de-
posited, namely: 1) a primary simatic
Crust, possibly analogous to modern
‘oceanic crust or island-arc-trench
domains (Bass, 1961; Folinsbee er a,
1968; Green and Baadsgaard, 1971
Ermanovics, 1973; Hubregise, 1976,
Wilson et at, 1974), and 2) an acid
igneous-metamorphic sialic basement
(Donaldson and Jackson, 1965; Ayres,
1974; McGlyan and Henderson, 1970,
Bel, 1971; Frith and Doig, 1975; Hender-
son, 1975, Baragar and McGlynn, 1976).
Inthe last reference, Baragar and
‘McGlynn assemble an impressive body
‘of evidence wnich suggests that denu-
ation of granitic rocks has taken place
‘concomitantly withthe volcanic evou-
tion of Keewatin (Abitibi) and Yellow-
knife greenstone sequences. Their
‘evidence and arguments include: 1)
observation or inferred existence of
basal unconformities: 2) documientation
of granite-derived clastic sediments;
‘and 3) the isotopic dating, in places, of
about three by. old granites. This
evidences usedaas the basis of an
‘Archean crustal model central which
is proposed existence of a continuous
sialic crust prior to the evolution of
(greenstone belts (Baragar and
McGlynn, 1976).
In view of the longtime span occupied
by the Archean era (defined here as 4.0
1026 by. ago), the possibilty hat
significant secular changes have oc-
curred during this ime, and diferences
in the depth of erosion and the crustal
level exposed - every terrain represents
but a segment intime and space and a
synthesis of all available information is
required for any model of Archean
crustal evolution. Supporters of an
original simatic crust point out the
abundance of matic and ultramafic
voleanic xenoliths within the isotopically
‘oldest orthogneisses in southwestern
Greenland, Labrador, Minnesota,
Rhodesia, Swaziland, India and Western
Australia (Viljoen and Viljoen, 1969,
Anhaeusser, 1973, Gikson, 1971, 1972,
1976; Naqvi, 1976). In contrast, sup~
porters of the sialic basernent hypothe:
sis point out occurrences of granitic
clasts within greenstone sequences
(Hunter, 1974; E.G. Nisbet, pers
commun, 1976; Baragar and McGiynn,
1976). Of central significance to this
problem is the observation of @ concom:
itant development of acid plutonic
activity and mafic-ultramatic volcanic
activity during atleast parts of the
‘Archean era (Fig 1). Thus, the lithology
ol the oldest stratigraphic unit, igneous
body, xenolith or clastic fragment in any
single terrain doesnt initselt necessar-
ily constitute evidence for the composi
tion of the eariiest crust inthis region ~
less ofall world-wide This truismis
‘evident with reference to modern tec:
tonic environments. For example, ocour-
rences of granitic rocks in island arcs
(eg, Gil, 1970) or mid-ocean ridges
(Coleman and Peterman, 1975; Engel
and Fisher, 1975) are no more in
evidence for an underlying sial than are
continental lood basalts (which may
include low-K tholaites) for an underly-
ing sima.Each Archean terrain, however, con-
ins the record of a succession of
‘events which effected a transition from
lone tectonic environment to another -
signitying a trend of crustal evolution,
These developments may be diachro-
nous (see Fig. 1) - resulting in a spatial
and temporal overiap of diferent stages
in diferent areas, For example, regions
inwhich an advanced stage of cratoni-
zation was reached may coexist with
adjacent regimes where an older crustal
‘segment has been litle modified by
+ i
Pan
eo
$4 =
e
Na granite nucleation L greenstone belt development
+
|
|
|
+
e
+
®e
etanm ante
impacts
Figure 1
Interpreted a'strbution of major events inthe
‘Archean evolution of southwestern Greer
lana (A), Nain Province (Labrador (6),
Rhodesian craton (C), Kaapyaal craton 0).
Pivara craton (E) and the Eastern Goiatetds
Province of the Vigarn craton (F). Stars -
Impact events: sa ines - ealy greenstone
Crust (utamatic-maic voteanics, minor
dacites and sediments) circled crosses -
tonalte-granodtorite suite; amonds and
arrows ~ minimum ages of late greenstones;
crosses - adamolite and quartz monzonte.
Other events - fr example, metamorphism,
‘minor igneous activty and sedimentation =
{are not shown on this oiagram.
younger tectonic and thermal events.
“The sum-total ofthe individual evolu-
tionary trends in different parts of the
‘Archean Earth must retlect an overall
though diachronous, trend of crustal
development. The central question in
this regard is whether greenstone belts
developed as: 1) intrasialic depressions;
Figure 2
Alternative models of Archean greenstone
bet evousion
Medel - Develogment of greenstone belts
as ntasiaic basins - ie, overying@ granitic:
‘metamorphic basement - accompanied by
‘anatexis in downbuekledifracrusta root
zones. Determation isassociated wih the
‘ascent of putons.
Madel2 - Qevelogment of greenstone belts
above oceanic crus! formedin a dvergent
Crustal gap o! ne Rea Sea type. iftng o! the
basic crust and ts partial meting at ower
levels results in Na-rich intermediate to acid
‘magma. Concomitant meting in subjacent
‘mantle diapirs beneath the ited zones
‘resutsin further matic-utramaticvoleanic
‘activity, Deformation of the supracrustal
racks results from the ascent of granitic
plutons.
2) simatic rt zones between diverging
sialic plates; or 3) by progressive
‘nucleation of granitic bathoiihs in
ensimatic regimes - the three alterna
tives are portrayed in Figure 2 Itis also
possible that two or all three interpreta-
tions are applicable to diferent green-
stone belts, Each of these models is
Model’3- Develonment of greenstone dels
byriting and! or downbucking ofa simatic
‘crust, represented by early greenstone
‘assemblages. Partial meting in crustal ot
zones and underying mane diapirsresuts in
foymaion of Na-sich granites and matic:
ultamatc volcanics respectively - the ater
etined as late greenstones. The sedimen:
tary units result rom erosion of acid volcanic
piles and of adjacent cratons - the result of
‘pit and stabilization o yet older gran
‘greenstone systems. Inthis model the sic
‘rust evolves in stages by the nucleation of
vastich grantes as a progressive and
iachronous process.
‘Symbois:gs -greenstones, sv - sediments
‘and acid volcanics: ap acig plutons hg ~
high-grade zone, mad - mante dap °g -
‘early greenstones' i - late greenstones:
‘un unconformity, az - anatectc zone,Geoscience Canada, Volume 5, Number 1
consistent with an occurrence of grani-
ticrocks at relatively early stages of
‘greenstone belt evolution. In the follow-
ing sections, the evidence bearing on
these alternatives is examined
Field Evidence
‘Arkosic conglomerates occur below
‘greenstones of the Hays River Group
(Oxford and Gods Lakes, Manitoba
Campbell et al, 1972), at low stratigra-
phic levels ofthe Abitibi Group (Holubec,
1972), in the Miminiska Group (Fort
Hope, Ontario) and in other localities
(Baragar and McGiynn, 1976). general
upwards increase in the importance of
‘ranite-derived sediments suggests
progressive uplft and denudation
of the granites.
Itis less clear, however, whether the
‘exposed granites represented inlers of
‘a continuous sialic basement, as sug-
gested by the frst model (Fig. 2) or
alternatively, outcrops of spatially separ-
ated sialic plates or nuclei, as suggested
by the second and third models, espec-
tively, The two latter models alow for
local greenstone-granite unconformities
where volcanic activity overlapped
peripheral zones of sialic plates or
ruclel. Clearly, however, nterms of the
‘second and third models the bulk of the
volcanic activity occurred within simatic
environments, ie. above newly formed
‘oceanic cust or older greenstones,
respectively, whereas sial basement.
greenstone unconformities would be
relatively rare. How extensive is the
evidence for basal unconformities and
granite basement outcrops underneath
Keewatin (Abitibi) and Yellowknife vol-
canic sequences? A review suggests,
that, in most instances, such occur-
rences have been inferred from:
1) Occurrences of granite-derived
sediments and cross-bedded quartzite
within, or atthe base, of volcanic
sequences.
2) Differential distribution of basic dykes
namely, their denser occurrence within
granites than within adjacent green-
stones. These relations were taken as
evidence for an older age of the granites,
suggesting that the dykes were feeders
ofthe volcanic flows (e.g, Heywoodand
Davidson, 1969)
3) The structural complexity of granitic
batholits is sometimes regarded as
evidence for their older age relative to
less deformed greenstones.
4) An absence in places of contact
‘metamorphic aureotes along granite:
{greenstone contacts is sometimes
regarded as an indication ofa relatively
younger age of the greenstones.
is suggested below that none of the
above observations is necessarily im:
plicit of a granitic basement. As pointed
ut botore, granite-derived sediments
could be derived from neighbouring
sialic plates or granitic nuciei. The
abundance of dykes within granitic
terrains is to large extent controlled by
the highly fractured nature of these
rocks, as contrasted to the ductile
layered nature of the supracrustal
greenstones. The Pilbara granite-
{greenstone terrain in Western Australia
Contains clear examples of Proterozoic
dykes which intrude granites but fail to
trangress granite-greenstone boun-
daries (Fig. 3). Noris the third criterion
Figures
‘An aerial photograph showing the reations
between a basic dyke, granite and green-
stones nan area north of Marbie Bar, Pibara
region, Western Australia. The area shown is:
‘about 19km across. Symbols: Oyke,
listed above implicit of granite-
‘greenstones age relations: inherent in
the development of composite plutonic
bodies is the progressive intrusion of
magmatic increments into semi-
consolidated granite - a process asso-
lated with syn- to late-magmatic
deformation, development of foldingand
neissosity. Superposed intrabatholithic
Structures cannot therefore be cons:
dered evidence for an older age relative
t0 less-detormed adjacent greenstones.
The structure ofthe latter is clearly
determined the external configuration
‘and mode of emplacement of the
batholiths (Anhaeusser of a/, 1969;
Hickman, 1975), bearing tle elation to
the internal endemic features of these
Plutons. An absence of thermal meta-
‘morphic aureoles along some grarite-
‘greenstone contacts does notin itself
‘suggest unconformable relations be-
‘9-grantte; vb - basic voleanies: va - acid
volcanics; T - Talga River. C- Coongan
Fiver, Arrow indicates the area where the