Life on Earth has been modifying the environment for billions of years. Green-plant photosynthesis was essential for the development of our current oxygen-rich atmosphere. The history of increasing oxygen in the atmosphere and ocean is complex, however, and significant free oxygen has been available in the atmosphere only during the past 2.2 billion years. Now new measurements by University of Rochester geochemists have uncovered evidence that even after 2.2 billion years ago, the amount of oxygen in the oceans remained low, perhaps up to the time when multicelled life began to proliferate a few hundred million years ago. Their work, published this week in the journal Science, has been supported in part by the NASA Astrobiology Institute, as well as other grants from NSF and NASA. Their paper is “Molybdenum isotope evidence for widespread anoxia in mid-Proterozoic oceans,” by G.L. Arnold, A.D. Anbar, J. Barling and W.T. Lyons.

The new evidence for a relatively anoxic (oxygen-free) ocean during the Proterozoic period has several implications for astrobiology. It warns us that additional information is needed to determine the actual concentrations of oxygen in the atmosphere and ocean during the Proterozoic. It suggests the possibility that the relatively recent in rise in oxygen in the ocean might have been an important environmental stimulus for the evolution of multicelled life. And the work provides important input to efforts to determine what signature of life could be detected in the atmospheres of planets circling other stars. To interpret atmospheric spectra of extrasolar planets, we need to understand how atmospheric oxygen content relates to the evolution of the biosphere. Astrobiologists are particularly interested in possible atmospheric signatures of microbial life, since the Earth has been a “microbe only” planet until relatively recently.

As noted in their press release, the research team has pioneered a new method that reveals how ocean oxygen might have changed globally. Previously, geochemists developed ways to detect signs of ancient oxygen in particular areas, but not in the Earth’s oceans as a whole. “This is the best direct evidence that the global oceans had less oxygen during that time,” says Gail Arnold, a doctoral student of earth and environmental sciences at the University of Rochester and lead author of the research.

Arnold examined rocks from northern Australia that were at the floor of the ocean over a billion years ago, using the new method developed by her and her coauthors. Their instrument — called a Multiple Collector Inductively Coupled Plasma Mass Spectrometer — was used to examine the chemistry of molybdenum’s isotopes within the rocks. Molybdenum is an element that enters the oceans through river runoff, dissolves in seawater, and can stay dissolved for hundreds of thousands of years. By staying in solution so long, molybdenum mixes well throughout the oceans, making it an excellent global indicator. The research team learned that the chemical behavior of molybdenum’s isotopes in sediments is different depending on the amount of oxygen in the overlying waters, and as a result that the chemistry of molybdenum isotopes in the global oceans depends on how much seawater is oxygen-poor. Compared to modern samples, their measurements of ancient rocks from Australia point to oceans with much less oxygen.

Their press release notes that “how much less oxygen” is the next question. A world full of anoxic oceans could have serious consequences for evolution. Eukaryotes, the kind of cells that make up all organisms except bacteria, appear in the geologic record as early as 2.7 billion years ago, but multicelled eukaryotes did not appear until much later. One of the paper’s authors, Ariel Anbar of the University of Rochester, previously suggested (with paleontologist Andrew Knoll of Harvard University) that an extended period of anoxic oceans might be the key to why the more complex eukaryotes barely eked out a living while their prolific bacterial cousins thrived.

“It’s remarkable that we know so little about the history of our own planet’s oceans,” says Anbar. “Whether or not there was oxygen in the oceans is a really straightforward chemical question that you’d think would be easy to answer. It shows just how hard it is to tease information from the rock record and how much more there is for us to learn about our origins.”

To help place the new work in context, Anbar addressed the question of whether the new work is consistent with previous estimates of oxygen in the Proterozoic atmosphere and oceans. He noted that the major lines of evidence usually cited for a rise in atmospheric oxygen from “almost nothing” to “something” about 2.2 billion years ago are:

a) The cessation of banded iron formation (BIF) deposition in the oceans

b) The appearance of terrestrial redbeds

c) The disappearance of easily oxidized minerals deposited in terrestrial (land) environments

d) The disappearance of the mass-independent sulfur isotope signature in marine sediments.

Of these, the last three all deal with oxygen in the atmosphere, so only the first (the BIF interpretation) is potentially contradictory in terms of the amount of oxygen in the oceans. To end BIF deposition, you need to change ocean chemistry such that the amount of iron dissolved in the oceans, and hence available to make BIFs, falls markedly. But there are at least two ways to accomplish this: through changes in either available oxygen or iron sulfides. Anbar notes that Don Canfield has suggested that the initial rise of atmospheric oxygen led to an increase in sulfate supply to the oceans, and that sulfate-reducing microbes turned the oceans sulfidic for a billion years. It is also possible that the BIF were deposited in very narrow windows of time. In this case, they may reflect temporary conditions triggered by large volcanic eruptions, rather than indicating the average state of the oceans over this billion-year time span.

We can also ask how the oceans could be in contact with an oxygen atmosphere for a billion years without themselves becoming oxygen-rich. Anbar notes that we have one similar analog today in the Black Sea, which is anoxic in spite of our oxygen-rich atmosphere. The oxygen content of the oceans is not a story of equilibrium with the atmosphere. The oxygen content is balanced by supply, mostly via equilibrium at the surface and physical mixing to depth, vs. loss, mostly due to the biota, consuming oxygen in the course of aerobic respiration. That loss rate of oxygen is largely dictated by the supply of organic carbon from the surface ocean, because it is during the aerobic respiration of that organic carbon that oxygen is consumed.

In the Black Sea, the deep waters are anoxic because of a confluence of two factors: Primary production in the surface oceans which supplies organic carbon to depth (organic remains settle toward the bottom), combined with very sluggish mixing of the system, which inhibits the supply of oxygenated surface water to depth. In the Proterozoic, Anbar thinks it is reasonable to postulate that the marine biota were comparably productive. With an atmosphere having only a few percent oxygen, this microbial activity might have kept the oceans anoxic. Thus the absence of oxygen in the ancient oceans is not evidence for low levels of biological activity, but might actually be in part the product of an active anaerobic biota in the oceans. These active microbes may have been mostly bacteria and archaea, however, not the eukaryotes that play such an important role in Earth’s life today.

This research suggests that we have much to do to determine the history of oxygen in Earth’s atmosphere and ocean. The Proterozoic seems to be a key period in the history of Earth’s biota. This is when the atmosphere was (perhaps only very gradually) changing to an oxygen-rich composition. It is also the time when life was preparing for the “Cambrian explosion,” when multicelled life forms and complex body types became common. Because the fossil record of the Proterozoic is so sparse, scientists know relatively little about the course of evolution during this span of more than one billion years. As we learn more about this crucial part of Earth’s history, astrobiologists will also be better able to identify ways we can look for evidence of inhabited planets around other stars.

I thank Ariel Anbar for providing a copy of this paper and commenting on its significance.