Precambrian Research 103 (2000) 101–124 www.elsevier.com/locate/precamres Vestiges of life in the oldest Greenland rocks? A review of early Archean geology in the Godtha˚bsfjord region, and reappraisal of field evidence for \3850 Ma life on Akilia John S. Myers *, James L. Crowley Department of Earth Sciences, Memorial Uni6ersity of Newfoundland, St John’s, Newfoundland, Canada A1B 3X5 Received 30 November 1999; accepted 18 May 2000 Abstract The Godtha˚bsfjord region of West Greenland contains the most extensive, best exposed and most intensely studied early Archean rocks on Earth. A geological record has been described of numerous magmatic events between 3.9 and 3.6 Ga, and evidence of life at \3.85 Ga and 3.8–3.7 Ga has been proposed from two widely-separated localities. Some of these claims have recently been questioned, and the nature of the best preserved remnants of the oldest known terrestrial volcanic and sedimentary rocks in the Isua greenstone belt are being reinvestigated and substantially reinterpreted. The first part of this article reviews the evolution of geological research and interpreta- tions, outlining the techniques by which the geological history has been determined and the ensuing controversies. The second part re-examines crucial field evidence upon which the antiquity of the oldest terrestrial life is claimed from the island of Akilia. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Early Archean; Gneiss complex; Geochronology; Oldest life; Greenland 1. Introduction sedimentary rocks during the late Archean. These rocks were repeatedly deformed and metamor- The Godtha˚bsfjord region lies in the centre of the phosed, developing gneissosity that was folded, refolded and transposed into new tectonic layering, Archean gneiss complex on the west coast of and recrystallizing during and after deformation at Greenland in the vicinity of Nuuk (Fig. 1, inset amphibolite or granulite facies (Bridgwater et al., map). Most of the Archean gneiss complex is 1976). derived from sheets of tonalite and granodiorite, In a belt 25–75 km wide, extending for 200 km and a small amount of granite and diorite, that were through Godtha˚bsfjord (Fig. 1), similar late intruded into basaltic volcanic rocks and minor Archean plutonic rocks were intruded into, and tectonically interleaved with, early Archean * Corresponding author. Tel.: +1-709-7378417; fax: +1- tonalitic and granodioritic gneisses, and both early 709-7372589. and middle Archean metavolcanic and metasedi- E-mail address: [email protected] (J.S. Myers). mentary schists and gneisses. 0301-9268/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0301-9268(00)00089-9 102 J.S. Myers, J.L. Crowley / Precambrian Research 103 (2000) 101–124 The first part of this article (Sections 2–4) vious observations and interpretations, and out- reviews knowledge of the early Archean rocks lining current controversies. Most of the early that formed before 3600 Ma, synthesizing pre- Archean rocks were intensely, and repeatedly, de- Fig. 1. Geologic map of the Godtha˚bsfjord region, compiled mainly from maps by Allaart (1982), Chadwick and Coe (1983, 1988), Garde (1987, 1989) and McGregor (1984), with terrane boundaries and Ikkattoq gneiss from McGregor et al. (1991) and geochronology from Nutman et al. (1996) and Nutman (1997). The inset map of Greenland locates Nuuk and the main regions of Archean gneiss that escaped pervasive ductile Proterozoic deformation. J.S. Myers, J.L. Crowley / Precambrian Research 103 (2000) 101–124 103 formed and metamorphosed at high grade during rocks cut by undeformed dykes from Archean late Archean events and, except in the vicinity of rocks and dykes that were deformed and meta- Isua, the present structure and appearance largely morphosed during the early Proterozoic. reflect these late Archean events (cover of Precam- Berthelsen and McGregor were the first to use brian Research, this volume). Therefore a brief basic dykes to distinguish between Archean tec- outline of the late Archean tectonic evolution is tonic, magmatic and metamorphic events in included in this review (Section 5), indicating cur- Greenland. Berthelsen (1955) recognized that ba- rent interpretations of how the early Archean sic dykes could be used to distinguish between rocks were intruded by, and tectonically inter- older and younger geological events in the south- leaved with, younger Archean rocks. eastern part of the island of Qila´ngaˆrsuit (Fig. 1). The second part of this article (Section 6) re-ex- McGregor (1966, 1968, 1973), during a detailed amines the field evidence from the island of Akilia investigation of the gneisses in the vicinity of upon which the antiquity of life is claimed to be Nuuk (formerly Godtha˚b and Nuˆk) between \3800 Ma (Mojzsis et al., 1996) and ]3850 Ma 1965–71, discovered that these basic dykes could (Nutman et al., 1997a; Mojzsis and Harrison, be used as a regional mapping criterion separating 2000), in the vicinity of the gneiss shown on the older from younger events. McGregor (1968) cover of Precambrian Research (this volume). called the dykes Ameralik dykes and recognized that intense deformation had severely modified the original structure of the dykes and their host 2. Pioneering studies in outer Godtha˚bsfjord rocks such that most dykes and host rocks were rotated into a new common gneissosity (e.g. cover Current understanding of the geology stems of Precambrian Research, this volume). Cross- from the pioneering work of Berthelsen (1955) cutting relationships were preserved only in small and McGregor (1966, 1968, 1973), who demon- areas of relatively low strain. Likewise, McGregor strated that basic dykes could be used as field (1973) described how the cross-cutting relation- criteria to subdivide the Archean gneisses into ships of granitoid rocks that were intruded after older and younger components. The basic dykes the dykes were also widely obliterated by intense cut across older rocks, structures and gneissosity, deformation that rotated most components of the and are cut by younger granitoid rocks that were gneiss complex into subparallel layers. subsequently deformed, metamorphosed and con- Many were sceptical of McGregor’s interpreta- verted to gneisses. tion, and some considered metamorphic grade to This technique had previously been developed be a key indicator of the relative age of rocks and in gneiss complexes elsewhere, especially in north- events. Based on a wider regional survey of the west Scotland by Peach et al. (1907) and Sutton Archean gneisses of West Greenland, Windley and Watson (1951), and in southern Finland by (1968, 1969) regarded the linear NNE-trending Sederholm (1923, 1926) and Wegmann (1931). belt of amphibolite facies gneisses in the Wegmann (1938) introduced this concept to Godtha˚bsfjord region, including the gneisses de- southern Greenland, distinguishing major tectonic scribed by McGregor, as derived by tectonic re- and metamorphic episodes (subsequently recog- working and retrogression of older granulite nized as Archean and Proterozoic) on the basis of facies gneisses to the north and south. He consid- intervening dyke swarms. During field studies be- ered that the Ameralik dykes of McGregor (1968) tween 1946 and 1955, Ramberg (1948) and Noe- cut older granulite facies gneisses to the south of Nygaard and Ramberg (1961), assisted by Ameralik (Fig. 1) and ‘can be followed… into the Berthelsen, applied this concept further north and linear belts where they become progressively de- presented the first fundamental subdivision of the formed and migmatised’ (Windley, 1969, p. 159). Precambrian gneiss complex from Godtha˚bsfjord Thus, in contradiction of McGregor’s hypothesis, northwards for 600 km. They used early Protero- the dykes and relatively straight NNE-trending zoic dykes to distinguish unmodified Archean belt of highly deformed amphibolite facies 104 J.S. Myers, J.L. Crowley / Precambrian Research 103 (2000) 101–124 gneisses through Godtha˚bsfjord were interpreted gneisses were highlighted by Coe et al. (1976) and as younger than regional granulite facies meta- Chadwick and Coe (1983) who discovered intra- morphism. However, McGregor’s hypothesis was Nuˆk dykes (Neriuneq and Qa´qatsiaq dykes) south subsequently supported by geochronology. Black of Ameralik (Fig. 1), broadly similar in appear- et al. (1971) discovered that the whole rock RbSr ance to many Ameralik dykes. Chadwick (1981) and PbPb isochron ages of gneisses that were cut described another, smaller and more controversial by Ameralik dykes in the Godtha˚bsfjord belt were group of dykes he called Tinissaq dykes, thought one billion years older than the widespread gran- to be intruded into an early phase of Nuˆk ulite facies metamorphism, and were far older gneisses. On the island of Qila´ngaˆrsuit to the west than any other terrestrial rocks known at that (Fig. 1), Chadwick (1981) recognized eight types time. of Ameralik dykes intruded into Amıˆtsoq gneisses Black et al. (1971) called the rocks that were and two types of amphibolite dykes, ‘similar to older than the Ameralik dykes ‘Amıˆtsoq gneisses’ two main categories of Ameralik dykes’ (ib. cit. P. and discovered that they were characterized by 221) cutting the Malene supracrustals. Chadwick extremely unradiogenic whole-rock lead isotopes. (1981), pp. 223–224) concluded that ‘limited evi- They described the sequence of events following dence of dyke intersections suggests some Amera- the intrusion of the Ameralik dykes as: eruption lik dykes may be