PALEOSOLS AND THE EVOLUTION OF ATMOSPHERIC OXYGEN - A CRITICAL-REVIEW

A number of investigators have used chemical profiles of paleosols to reconstruct the evolution of atmospheric oxygen levels during the cour se of Earth history (Holland, 1984, 1994; Kirkham and Roscoe, 1993; Oh moto, 1996). Over the past decade Holland and his co-workers have exam ined reported paleosols from six localities that formed between 2.75 a nd 0.45 Ga. They have found that the chemical profiles of these paleos ols are consistent with a dramatic change in atmospheric P-O2 between 2.2 and 2.0 Ga from less than or equal to 0.002 to greater than or equ al to 0.03 atm (Holland, 1994). Ohmoto (1996) examined chemical data f rom twelve reported paleosols ranging in age from 2.9 to 1.8 Ga-1 He c oncluded that these chemical profiles indicate that atmospheric P-O2 h as not changed significantly during the past 3.0 Ga. We seek to resolv e the conflict between these reconstructions through broader examinati on of the paleosol literature, both to determine which reported paleos ols can be definitively identified as such and to determine what these definite paleosols tell us about atmospheric evolution. We here revie w reports describing over 50 proposed paleosols, all but two are older than 1.7 Ga. Our review indicates that 15 of these reported paleosols can be definitively identified as ancient soils. The behavior of iron during the formation of these 15 paleosols provides both qualitative and semiquantitative information about the evolution of the redox stat e of the atmosphere. Every definitely identified pre-2.44 Ga paleosol suffered significant Fe loss during weathering. This loss indicates th at atmospheric P-O2 was always less than about 5 x 10(-4) atm prior to 2.44 Ga. Analysis of the Hokkalampi paleosol (2.44-2.2 Ga) (Marmo, 19 92) and the Ville Marie paleosol (2.38-2.215 Ga) (Rainbird, Nesbitt, a nd Donaldson, 1990) yield ambiguous results regarding atmospheric P-O2 . Loss of Fe during the weathering of the 2.245 to 2.203 Ga Hekpoort p aleosol (Button, 1979) indicates that atmospheric P-O2 was less than 8 x 10(-4) atm shortly before 2.2 Ga. The presence of red beds immediat ely overlying the Hokkalampi, Ville Marie, and Hekpoort paleosols sugg ests that by about 2.2 Ga there was an unquantified but substantial am ount of oxygen in the atmosphere. Iron loss was negligible during form ation of the 2.2 to 2.0 Ga Wolhaarkop (Holland and Beukes, 1990) and D rakenstein (Wiggering and Beukes, 1990) paleosols and during formation of all the later paleosols we reviewed. Thus, atmospheric P-O2 probab ly has been greater than or equal to 0.03 atm since sometime between 2 .2 and 2.0 Ga.