Atmosphere of the early earth

No evidence for reducing conditions in ancient paleosols

Ask any student of historical geology what the Earth's atmosphere was like in the Precambrian and you will get some fairly predictable responses. Oxygen was "definitely" not as plentiful as today because, it is said, the Earth's biosphere (which maintains the oxygen cycle) was nothing like that of today. In the Precambrian, plants which consumed carbon dioxide and released oxygen were represented only by single celled algae. The earlier you look in the Precambrian, the less evidence there is for free oxygen. In the Archaean, pre about two thousand five hundred million years ago (pre 2.5 Ga), it has been said that the atmosphere was either neutral (a mixture of nitrogen, carbon dioxide, water vapour and perhaps a little hydrogen) or reducing (with ammonia, methane, carbon dioxide, water vapour and perhaps hydrogen). The reducing atmosphere is thought to be the original Earth's atmosphere by analogy: the outer planets (Jupiter, Saturn, Uranus, Neptune) have reducing atmospheres and this has long been thought also to represent the "original" state of the Earth's atmosphere.

This "evolution of the atmosphere" scenario is of great importance for theories of abiogenesis. Miller, in his classic (1951) experiments, produced amino acids, "the building blocks of life", using a reducing atmosphere to provide the raw materials. Miller's experiments continue to be cited because there is no other viable source of amino acids from which to construct proteins and other complex organic chemicals. However, a neutral atmosphere results in severely impoverished reaction products, whereas an oxidising atmosphere has no amino acid production of any significance.

Overviews of research on the Earth's early atmosphere have been published at regular intervals. Direct evidences of a reducing atmosphere have been claimed in the past, but these evidences are no longer regarded as conclusive. There is a growing consensus that in the earliest period for which data is available, the atmosphere was neutral, with negligible amounts of free oxygen. Such is the conclusion, for example, of Kasting (1993). Prior to 2.0 Ga, although there is no valid evidence for a reducing atmosphere, whatever oxygen was around is believed to have been consumed in the oxidation of other materials: organic matter, iron bearing minerals and volcanic gases. Estimates of oxygen levels are about 10-13 of the present atmospheric levels (PAL). One of the best evidences for the low levels of oxygen (giving a neutral atmosphere) comes from the loss of iron from pre-2.2 Ga paleosols (fossil soils). After about 2.0 Ga, oxygen levels increased significantly to about 1.5% of PAL oxygen. So the consensus in the early 1990s has been:

Early Archaean (pre 3.0 Ga)	Reducing atmosphere (but unsupported by data)
Late Archaean (3.0 - 2.5 Ga)	Neutral atmosphere with some free oxygen
Proterozoic (2.5 - 0.6 Ga)	Oxidising atmosphere with about 1.5% PAL oxygen

In a major reexamination of the paleosol evidence, Ohmoto (1996) has effectively challenged the concept of a neutral atmosphere. He argues that the minimum oxygen pressure for the 3.0 - 2.2 Ga (for which paleosol data is available) is about 1.5% of the present level. We now turn to look more closely at the new data and Ohmoto's analysis.

The arguments are based on the occurrence of compounds of iron in certain sedimentary rocks. Iron in the ferrous state (Fe²⁺) can dissolve relatively easily in oxygen-free water, but is converted to the insoluble ferric state (Fe³⁺) in an oxidising environment. Previous studies of certain Precambrian rocks identified as weathered horizons (paleosols) have suggested a general loss of iron, which has been interpreted as evidence for either a neutral or a reducing atmosphere.

Ohmoto's research was stimulated by some apparent anomalies in the conventional analysis. He found that not all paleosol sections of >2.2 Ga showed iron loss. Even in sections that did show Fe loss, only a minority of samples were depleted in iron. Furthermore, many of the post 2.2 Ga paleosols had lost iron (and in such cases, an atmosphere with some free oxygen is accepted). To resolve the numerous questions raised by this data, Ohmoto gathered new data involving more detailed measurements of atomic ratios. This enabled him to obtain a chemical signature which could be associated with weathering in reducing/oxidising/other environments.

Problems with the conventional interpretation

Theoretically, the lower the level of oxygen ions in water (and the greater the level of hydrogen ions), the more iron can go into solution. However, the rates of dissolution have to be assessed experimentally. It is found that the reactions for Fe³⁺ compounds proceed very slowly. The prediction is that Fe²⁺ will be lost more readily than Fe³⁺. Using Titanium as a `standard' immobile element, the prediction is that a "reduced"-type (R- type) paleosol will have significant reductions in the ratio Fe²⁺/Ti but little or no decrease in Fe³⁺/Ti. According to Ohmoto, none of the paleosol sections examined yielded this characteristic. Thus, there are no paleosols that support the idea that the earth's early atmosphere was reducing (or neutral, for the same reasons).

A new interpretation of the paleosols

Ohmoto's research allowed him to classify the observed paleosols according to their isotopic characteristics. The details of this need not concern us in this review, but the conclusions are of considerable interest.

(a) Oxidised (O-type) paleosols The four isotopic trends characterising the formation of soils under oxic conditions today are retained in the Precambrian paleosols.

(b) Hydrothermally-altered (H-type) paleosols Isotopic trends in these cases suggest that these rock sections (and possibly other paleosols) are not paleosols at all, but hydrothermally-altered rocks. Fluids carrying ions have passed through these rocks and imparted a chemical signature to them.

(c) Mixed processes (M-type) paleosols These rocks have isotopic fingerprints which can be related to both O-type and H-type paleosols. Ohmoto says: "These characteristics suggest that M-type paleosols formed under an oxic atmosphere, but their Fe[iron] chemistry was modified during and/or after soil formation." Ohmoto also notes that M-type paleosols are the most common of the three types.

Implications of the new interpretation

Having discussed the three types of paleosol observed and possible modes of formation, Ohmoto concludes that the minimum pressure of atmospheric oxygen consistent with the data is greater than 1.5% PAL for the entire period of 3.0 - 1.8 Ga.

This new analysis puts increasing pressure on all "reducing atmosphere" interpretations of the Earth's early atmosphere. There is no observed trend of reducing -> neutral -> oxidising. As far as data is concerned, the Earth's atmosphere has always been oxidising. Theories of abiogenesis which require a reducing atmosphere are pushed further into a realm of speculation supported by theoretical models but not by empirical data.

An atmosphere with free oxygen points to the contemporaneity of plants which photosynthesise. However, to date, studies of organic life in the Archaean have suggested the existence of only bacteria and single-celled algae. But this evidence is not plentiful. Even the growth mounds, the Precambrian stromatolites, now appears to be better explained as having an abiotic origin (Grotzinger & Rothman, 1996). This meagre fossil evidence has allowed the earlier postulate of a neutral atmosphere during the later Archaean. With the new evidence of a significantly oxygenated atmosphere, it may be inferred that more plant life was around than we have direct evidence for. Ohmoto says: "Terrestrial biomass on the early continents may have been more extensive than previously recognised". Following through this thought leads to a number of interesting possibilities for further research:
(i) with more rigorous investigation, will body fossils of this more extensive biomass be found?
(ii) were post-deposition processes less conducive to fossil formation in the PreCambrian than in the Phanerozoic?
(iii) were sedimentary processes less conducive to fossil formation in the PreCambrian than in the Phanerozoic?

Grotzinger, J.P. & Rothman, D.H. 1996. An abiotic model for stromatolite morphogenesis, Nature, 383(3 October), 423-425.
Kasting, J.F. 1993. Earth's early atmosphere. Science, 259, 920-926.
Ohmoto, H. 1996. Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota. Geology, 24(12), 1135-1138.

David J. Tyler (May 1997)

Return to top of page