Microfossil-like objects from the Archaean of Greenland: A cautionary note

Abstract

Recent reports have described 'yeast-like microfossils' (Isuasphaera isua Pflug) in 3,800-million year old metaquartzites from the Isua supracrustal belt of south-west Greenland. A biogenic interpretation of these objects is inconsistent with the tectonic history of the Isua region, with the petrology of the metaquartzites, and with the morphology of the microstructures themselves. The putative microfossils are indistinguishable from limonite-stained fluid inclusions: microstructures which are demonstrably inorganic and post-depositional in origin. As such, it is contended that these objects should not be regarded as evidence of early Archaean life forms.

Full text

Nature
Vol.
289
1/8
January
19151
Microfossil-like objects from
the Archaean
of
Greenland: a cautionary note
D. Bridgwater*~,
J.
H. Allaart*,
J.
W.
Schopft,
C.
Kleint
#,
M.
R. Waltert**, E.
S.
Barghoornt,
P. Strothert,
A.H.
Knoll§ & B. E. Gormanll
* Geological Survey of Greenland
0ster
Voldgade 10,
DK-1350
Copenhagen K,
Denmark
t Precambrian Paleobiology Research Group,
Department
of
Earth
and Space Science, University of California, Los Angeles, California
91124
t
Department
of Biology, Harvard University, Cambridge,
Massachusetts 02138
§ Department of Geology, Oberlin College, Oberlin, Ohio 44074
11
Department
of Geology, The University of Western Ontario,
London, Canada
N6A
5B7
Recent reports
14
have described "yeast-like microfossils"
(Isuasphaera isua Pflug2)
in
3,800-Myr old5 metaquartzites
from the Isua supracrustal belt of south-west Greenland. A
biogenic interpretation of these objects
is
inconsistent with the
tectonic history of the Isua
region-,
with the petrology of the
metaquartzites, and with the morphology of the microstructures
themselves. The putative microfossils
are
indistinguishable from
limonite-stained fluid inclusions: microstructures which are
demonstrably inorganic and post-depositional
in
origin. As
such, we contend here that these objects should not be regarded
as
evidence of early Archaean life forms.
The geological history of the Isua supracrustal belt
is
sum-
marized
in
Table
1.
Several episodes of deformation, geo-
chemical alteration and amphibolite-facies metamorphism have
been recognized,
in
sharp contrast to slightly younger Archaean
strata from which stromatolites9•
10
and apparently bona fide
microfossils11-
11
have been reported.
Evidence of penetrative deformation
is
seen throughout the
Isua succession where all lithologies record high strains. Few
primary sedimentary and igneous structures are preserved apart
from the gross lithological layering which allows a rudimentary
stratigraphy to be described. Even
in
fold hinges or inclusions of
competent material
in
a ductile matrix where strain
is
compara-
tively low a strong linear fabric
is
developed and primary
features are only recognizable on surfaces normal to the maxi-
mum extension. Original discordant structures, such
as
the
contacts between the supracrustal rocks and the intrusive
Amitsoq gneisses have been rotated to near-parallel. The
absence of strain markers
in
the main quartzite horizon (from
which the fossil-like microstructures were found), obscures
evidence of deformation. Nonetheless, thin layers of magnetite
and cummingtonite
in
the quartzites exhibit extension fabrics
similar to those
in
neighbouring ironstones. There
is
no evidence
that the quartzite horizon escaped the episode(s) of high
deformation.
Large-scale geochemical alteration
is
inferred from rocks
interpreted
as
volcanic
in
origin. Basic rocks record massive
remobilization of alkalis, silica and
CO
2 (ref. 14), whereas acid
rocks are strongly enriched
in
K2O (refs 6, 7)
or
carbonate
minerals. Ultrabasic units (dunites and peridotites) are com-
monly transformed into schistose assemblages of talc, Mg-
amphibole and carbonates. Sedimentary carbonate units display
high initial Sr-isotope ratios (see
W.
Compston
in
ref. 15)
attributed to massive metasomatic changes
in
constituents.
11
Present address:
The
Geological Museum,
0ster
Voldgade
5,
DK-1350
Copenhagen.
# Permanent address: Department of Geology, Indiana University, Bloomington, Indiana
47405.
••
Present address: Bureau of Mineral Resources, Geology and Geophysics, PO Box 378,
Canberra, ACT, Australia 2601.
0028--0836
/81/020051-03$01.00
51
At
least
five
regional metamorphic events occurred between
3,700 and 1,600 Myr (Table 1). The supracrustal rocks reached
amphibolite-facies metamorphism twice: once associated with
the emplacement of the Am1tsoq gneisses
(-3,700
Myr), and
again following the intrusion of the Ameralik dykes (before
3,000 Myr). Efforts to quantify temperatures and pressures
attending these metamorphic events have been hampered by the
Fig. 1 Optical photomicrographs showing recrystallized, equi-
granular quartz grains (in plane polarized light;
a)
and non-
biogenic, apparently limonite-stained microfossil-like objects (in
transmitted white light;
b-f)
in petrographic thin sections of
greenish-white quartzite from the 3,800 Myr Isua supracrustal belt
of Greenland;
a,
arrows point to fine actinolite needles;
b,
arrows
point to clear, ellipsoidal, fluid inclusions;
c,
d,
arrows point
to
limonite-stained quartz grain boundaries (also visible in b);
f,
arrow points to gas bubble within an elongate, mineral-stained
inclusion.
b-f
(scale as
in
b) are from thin section no. 60464,
Paleobotanical Collections, Harvard University, Cambridge,
Massachusetts.
© I 981 Macmillan Journals Ltd

© Nature Publishing Group1981
52
polymetamorphic nature of the rocks and their common retro-
gression to chlorite-bearing greenschist facies assemblages.
High-grade mineral assemblages recognized from different rock
types are listed
in
Table
2.
With specific reference to the pelitic
units, the assemblage biotite
+garnet+
quartz+
staurolite and
the presence of kyanite
16
are consistent with temperatures of
500-600
°C
and pressures of 200-500 MPa. Independent evi-
dence of the temperatures of metamorphism
is
given
by
the
oxygen isotope fractionation
in
quartz-magnetite pairs of the
main quartzite horizon
17
. These ratios indicate a mean (and
likely minimum) equilibration temperature of 390 °C, which
we
interpret
as
reflecting a late (post-3,000 Myr) event following
regional amphibolite-facies metamorphism.
The 'fossiliferous' rock, specimen no. 2377,
is
a relatively
coarse-grained, sugary, greenish-white quartzite. It
is
composed
of quartz (95%), actinolite (4%) and talc
(1
%). The quartz
is
equigranular and totally recrystallized, with most grains cluster-
ing
in
the 0.2-0.3 mm size class. Neither the presence of "chal-
cedony"
in
the rock1·3
.4,
nor its supposed "cherty" texture1·2
.4.
could be confirmed. Much of the actinolite
(Ca1.44Nao.04Mno.o7Feo.4s)
(Mg3_79Fe1.13Alo.01)
Sis023
is
of
a grain size similar to the quartz, and fine acicular needles can
be found within many individual quartz grains (Fig.
la).
The talc (Mgs.19Feo.d (Sh.90Alo.d022
is
seen as fine
"beards" along quartz grain boundaries. The mineralogy
of the quartzite does not contradict
our
conclusion that the
quartzite unit was subjected to amphibolite-facies regional
metamorphism.
Evidence of high grade metamorphism and extreme defor-
mation does not entirely preclude the survival of microfossils,
although the chances of preserving intact such delicate carbo-
naceous bodies as those described following complete recrystal-
lization of their host rock seem to be remote. The morphology of
the microfossil-like objects
in
thin section (including specimen
no. 2377 from which Isuasphaera was first described2) convinces
us
that a biogenic interpretation
is
unfounded. The Isua micro-
Table 1 Simplified table of events
in
the Isua area
Events Time (Myr
BP)
10. Resetting of
K-Ar
and
Rb-Sr
mineral ages
25
local
granite intrusion
26
and static thermal metamor-
phism to approximately 300°, faulting in green-
schist facies conditions. 1,600
9. Intrusion of
E-W,
N-S
and
NE-SW
Proterozoic
dykes.
2,100-
8. Transcurrent faulting juxtaposing Isua
supracrustals and Amisoq gneisses against Nuk
(-3,000
Myr) gneisses in
NW
of area. Partial
retrogression of amphibolite-facies assemblages
in
supracrustals.
7. Deformation with local strong shearing concen-
trated
in
supracrustal units. Amphibolite-facies
metamorphism and strong deformation of
Ameralik dykes within supracrustals.
3,000-
6. Intrusion of
E-W,
N-S
Ameralik dykes.
5. Retrogression
and
shearing particularly along
contact of supracrustal rocks. Intrusion of Amit-
soq pegmatites and leucogranites.
3,500-
4. Amphibolite-facies metamorphism, deformation
to produce major stretching fabric in supracrus-
tals.
3. Intrusion of Ami'tsoq tonalitic
and
granodioritic
gneisses.
3,700-
2.
Deformation of supracrustals. Fabric locally
preserved, mineral assemblages commonly
retrogressed during later events.
1.
Deposition of Isua sediments, volcanism, post-
depositional hydrothermal activity.
3,800-
Ages
quoted are rounded to the nearest 100 Myr. Based on published
and unpublished Rb-Sr,
Sm-Nd,
Pb-Pb
whole-rock,
Pb-Pb,
U/Pb,
Rb-Sr
and
K-Ar
mineral determinations. See refs
5-8
for original refs.
Nature
Vol.
289
1/8
January 1981
Table 2 Amphibolite-facies mineral assemblages* in the Isua
supracrustal rocks
Rock type
Ironstones (and Mg-rich analogues)
Basic volcanic rocks
Mg-rich and ultrabasic volcanic
rocks
Carbonate rocks and calc-silicates
Pelites
Assemblage
Quartz, magnetite,
grunerite / cummingtonite
Hornblende, plagioclase
(andesine-labradorite),
quartz, magnetite, sphene
(diopside)
Talc, anthophyllite, carbonate
minerals (quartz,
plagioclase)
Diopside, anthophyllite, Fe-
carbonate, calcite
Biotite, quartz, garnet,
magnetite (staurolite,
kyanite)
* Minerals
in
parentheses are only sporadically preserved.
fossil-like objects are reported to "consist predominantly of
carbon
...
characterized
by
a high degree of bitumunization"1
preserved "partly
in
a charred condition, partly close to a final
stage of graphitization"4; they have shapes ranging from ellip-
soidal to elongate with "individuals
...
found
in
the assemblages
which resemble a biscuit, a dumb-bell, or a pair of tears"1. The
objects range
in
size from 10 to 40 µm (with "mature cells" said
to occur towards the upper end of the size range3
),
and they are
described as being "hollow" with their interiors "partly filled
with organic matter or fine-grained pyrite" which
is
"dark
brownish"
in
colour1·3. They are also reported to exhibit struc-
tures interpreted
as
"cell walls", "cell vacuoles", "gas vacu-
oles", "multilaminate sheaths", "budding structures",
"bud
scars", "cross walls", "remnants of protoplasm" and evidence of
"filamental growth"l.3. By comparison with illustrations of
dried, carbol fuchsin-stained yeast cells, and on the basis of the
features noted above, these objects have been interpreted as
representing carbonaceous cellularly intact remnants of "yeast-
like organisms", possibly of
"an
evolutionary level which
is
far
below that of yeasts of today" but which "may represent a
half-way line between a microsphere-like protobiont and
subsequent evolution"1·3.
Our
thin sections contain microstructures which conform
in
size, shape and all other optically detectable properties to the
putative microfossils previously described, leaving no doubt that
our material
is
identical to that previously interpreted
as
bio-
genic. We find, however, that the populations of these objects
span a decidedly broader size-range than that previously repor-
ted: measurements of 250 such objects
(in
three thin sections)
show that they range
in
length between 2 and 115 µm, with a
mean length of
-23
µm (and a variance,
s2,
of 38.2).
Two principal categories of fossil-like objects occur in the thin
sections: (1) clear, spherical to ellipsoidal structures (Fig. 1 b)
occur only within quartz grains (rather than at grain boun-
daries); these structures are comparable to fluid inclusions:
inorganic, metamorphically produced micro-structures that are
of well known and relatively common occurrence
in
other,
similarly altered metasediments
18
;
(2)
yellowish to reddish-
brown (apparently limonite-stained) structures that exhibit a
continuum of shape from spheroidal to ellipsoidal to sub-
angular to completely euhedral (Fig.
lb-f).
With the apparent exception of the occurrence of "multi-
laminate sheaths" (of which
we
find no evidence) members of
this second category are identical to forms described as Isuas-
phaera 1
--4.
However, the population includes rhombohedral
(Fig.
le)
and obviously crystalline individuals (see also ref. 1 Fig.
3, parts 2, 4, 11, 12) which do not exhibit the pattern of
morphological consistency characteristic of fossilized micro-
organisms
19.
Furthermore the objects show a marked concen-
tration at quartz grain boundaries (Fig.
le,
d), a feature com-
monly observed
in
fluid inclusions
in
recrystallized granites
20

Nature
Vol.
289
1/8
January 1981
and not likely to be the result of a primary sedimentological
distribution. The reported "budding", which has previously
been attributed to a biological origin,
we
interpret as a 'necking-
off' process commonly occurring during recrystallization of
multiphase inclusions
21 ,
while the reported "gas vacuoles" are
bubbles trapped within the objects during their formation and
recrystallization
18
•
Other suggested biological features (bud
scars, cross walls, and so on) are more consistent with an
inorganic origin, for example, the staining of the surfaces of fluid
inclusions and negative crystals by secondarily mobilized
limonitic iron oxides which were also deposited along grain
boundaries. Many of the quartz grain boundaries are,
in
fact,
stained with limonite and are indistinguishable
in
colour and
texture from the coating that outlines the fossil-like objects (see
ref. 4, plate 2, Figs
2,
5, 6).
Thus consideration of the geological, petrological and
morphological evidence leads
us
to conclude that the fossil-like
microstructures
(Isuasphaera isua)
described from the 3,800-
Myr old Isua supracrustal rocks are
in
fact non-biogenic arte-
facts of inorganic, post-depositional origin. Their reported
"yeast-like" morphology and fine structure are inconsistent with
the tectonic history of the rocks
in
which they occur and are at
variance with the remainder of the Precambrain fossil record
as
now known
22
•
The occurrence of hydrocarbons within these
microstructures3,
if
confirmed, would be equally consistent
with both a biological and non-biological origin of these
compounds. We believe that the microstructures are non-
biogenic, and while graphitic carbonaceous matter occurs in
the Isua metasediments
4'
23
'
24
it remains to be established
whether this
is
biogenic
or
abiogenic
in
nature and to what
extent it may be of postdepositional rather than solely
syngenetic origin.
We thank
M.
Schidlowski for providing sample no. 2377 and
P.
Appel and C. Walters for additional material, and
W.
S.
Fyfe,
G. Perry,
R.
Dymek, R. Kerrich and J. D. Meloche for informa-
tion and discussion. Mapping at Isua (D.B. and J.H.A.)
is
part of
the programme of the Geological Survey of Greenland. Field
work was supported
by
NA
TO
research grant no. 949 and by a
NSERC (Canada) grant to
W.
S.
Fyfe (B.E.G.). We thank other
members of the NATO supported research group
8
for dis-
cussion. Laboratory work was supported by NSF grants 77-
22518 and 79-21777 and
by
NASA grants
NGR
05-007-407
and NSG 7489 (to J.W.S.);
by
NASA grant NGL 22-007-067
and
by
NSF grant
EAR
78-24237 (to E.S.B.)
by
NSF grant
EAR
76-117
40 (to C.K.).
Received 2 May, accepted 10 November 1980.
\.
Pflug, H. D. Naturwissenschaften 65,
611-615
(1978
).
2.
Pflug, H. D. Oberhess. Narurw. Z.
44,
131-145
(1978).
3. Pflug. H. D.
&
Jaeschke-Boyer, H. Nature 180,
483-486
(1979).
4. Pflug, H. D., Jaeschke-Boyer, H. & Sattler, E. L. Microsc.
Acta
82,
255-266
(1979).
S.
Moorbath, S., O'Nions,
R.
K.
&
Pankhurst, R.
J.
Nature 245,
138-139
(1973).
6. Bridgwater, D., Keto,
L., McGreaor, V. R.
&
Myers, J. S.
in
Geology
of
Grun/and
(eds
Escher. A.
&
Watt,
W.
S.)
18-75
(Geological Survey of Greenland, 1976).
7. Allaart, J. H .
in
The Early History
of
the Earth (ed. Windley,
B.
F
.)
177-189
(Wiley, New
York, 1976).
8. Bridgwater, D.
et
al
. Rapp. Gr/Inland geo/.
Undm.
95,
66-71
(1979).
9. Walter.
M.
R
.,
Buick, R.
&
Dunlop, J .
S.
R. Nature 284,
443-445
(1980).
10. Lowe, D. Nature
184,
441-443
(1980).
11. Schopf, J. W.
&
Barghoorn, E. S. Science 156,
508-S
12 (1967).
12. Muir, M. D.
&
Grant,
P.R.
in
The Early History
of
the Earth (ed. Windley, B. F.)
595-604
(Wiley, New York. 1976).
13. Knoll, A . H . & Barghoorn, E.
S.
Science
198,
396-398
(1977).
14. Gill , R. C. 0 ., Bridgwater, D.
&
Allaart, J. H . Spec. Pub/. geo/. Soc. Aust. (in the press).
IS. Apple,
P.
W. U. Precamb.
Res.
11, 83 (1980).
16
. Boak, J. L.
&
Dymek,
R.
F.
Geo/. Soc. Am. Abstr. (in the press
).
17. Perry, E. C., Ahmad,
S.
N.
&
Swulius, T.
M.
J.
Geo/.
86,
223-239
(1978).
18. Deicha, G.
Les
Lacunes
des
Cristaux
et
leurs
Inclusions
Fluides.
Signification dans
la
Genese
des Gftes Miniraux et des Roches (Masson et Cie. Paris, 1955).
19. Schopf, J.
W.
Origins
of
Life
7,
19-36
(1976).
20. Wilkins, R.
W.
T. & Barkas,
J.P.
Contr. Miner. Petrol. 65,
293-299
(1978).
21. Roeddcr, E. Fluid lnclusion Res.
Proc.
Coif,, 4 (1968).
22. Schopf, J. W. A.
Rm
Earth planet. Sci. 3,
213-249
(1975).
23. Nagy. B., Zumber11c , J. E.
&
Nagy,
L.A
.
Proc.
natn.
Acad
. Sci. U.S.A. 72,
1206-1209
(1975).
24. Schidlowski, M
.•
Appel.
P.
W.
U., Eichmann, R. & Junge, C. E. Geochim. cosmochim.
Acta
43,
189-199
(1979).
25. Pankhurst, R. J., Moorbath, S., Rex, D. C.
&
Turner, G. Earth planet. Sci. Lett. 20,
157-170
(1973).
26. Kalsbcck , F., Bridgwater, D. & Boak, J. Rapp. Grt1nlands geol. Unders.
100,
;J-75
(1980).
0028--0836/81/
020053-04$01.00
Amino acids and hydrocarbons
-3,800-Myr
old in the
Isua Rocks, southwestern Greenland
Bartholomew Nagy*, Michael H. Engel*;,
John E. Zumberge*§, Hiroshi Ogino*
&
Sai Y. Changtll
53
*
Laboratory of Organic Geochemistry, Department of Geosciences,
The University of Arizona, Tucson, Arizona 85721
t
Department of Pharmacology, The University of Arizona, Tucson,
Arizona 85721
The Isua supracrustal belt consists of metamorphic rocks,
including metamorphosed sediments, and are the oldest known
rocks on Earth.
It
has been postulated
1'2
that hydrocarbons
recently detected in these rocks
are
remnants of organisms that
lived -3,800 Myr ago. This popular interpretation
is
inconsis-
tent with the high-temperature history of the Isua rocks. Various
types of hydrocarbons were liberated from these rocks after
repeated solvent extraction by matrix dissolution
or
pyrolysis.
However, the Isua rocks have been metamorphosed to the
upper greenschist
or
amphibolite fades
(40~00
°C); the
metamorphic episodes lasted
>10
6
yr. Consequently, it
Is
very
important that the origins of the key organic constituents be
established. In the study reported here both hydrocarbons and
amino acids were detected in Isua rock samples. In addition to
the common biological amino acids, geologically unstable amino
acids and biologically uncommon ones were detected (such as
sarcosine). The extent of amino acid racemization suggests the
apparent continuous diffusion of biochemicals into the Isua
rocks from encrusting lichens since the end of the last
Ice
age.
Also, the cold-temperature history (which retards racemization)
suggests that the amino acids may be modern to a few tens of
thousands of years old. n-Alkanes lacked odd/even carbon
chain preference which may indicate some antiquity. However,
the hydrocarbons also resemble a recent petroleum distillate
fraction. Laboratory experiments, supported by kinetic and
thermodynamic calculations, show that amino acids and aroma-
tic and saturated aliphatic hydrocarbons, including pristane,
could not have survived the known metamorphic history of the
Isua rocks.
Samples from the Isua banded iron formation
(-150
km
north-east of Godthab) were analysed. Sample 155782 was an
interior portion of the banded ironstone
(-50
g)
that contained
a limonite-filled microcrack. Samples 3015, 3016, 3022 and
3028
(-50
g each) had weathered surfaces. Porosity, air and
water permeabilities of samples 3015 and 3016 were 2.3%,
1.0 x
10-
2
mdarcies and 1.2 x
10-
6
mdarcies; and 2.2%, 4.6 x
10-
1
mdarcies and 1.3 x
10-
1
mdarcies, respectively
3•
Isua
sample 3015 was less extensively weathered than Isua 3016 and
thus had a lower water permeability.
To
minimize laboratory contamination, all glassware was acid
cleaned, solvents and acids were redistilled from reagent
or
spectral grade, water was triple distilled in glass and so on4'5. The
rock surfaces were drilled
off
and the samples pulverized.
Excessive ball milling
(>20
h)
was avoided because it caused
some racemization of amino acids. The pulverized samples and
corresponding blanks were refluxed with water for 8
h.
The
filtrate residues were hydrolysed for 24 h at 100
°C
in
6 M HCI.
The samples were then desalted
by
ion exchange chromato-
graphy4·5
and the
N-PFP-(
+
)-2-butyl amino acid diastereomers
were prepared
4•5
and analysed for their
D/L
amino acid ratios
by
gas chromatography. Amino acids in the water extract of sample
3028 were analysed by combined gas chromatography-chemical
ionization mass spectrometry (GC-CIMS). Part of the water
extract of Isua sample 3016 was analysed
by
an amino acid
analyser.
Present addresses: iCarnegie Institution
of
Washington. Geophysical Laboratory, 2801 Upton
St. NW, Washin11ton,
DC
20008;
§Cities Service Co., Box 50408, Tulsa, Oklahoma 74150;
11
McNeil Laboratories, 500 Office
Center
Drive, Fort Washington, Pennsylvania 19034.
©
1981 Macmillan Journals Ltd