For purposes of illustration, three distinct stages in our biosphere's evolution are envisioned. The first stage is the interval between the origin of life the the development of oxygenic photosynthesis. The atmosphere was at least mildly reducing and was dominated by CO₂, N₂, H₂ and H₂O. The modest amounts of O₂ created by the photodissociation of water vapor and loss of H₂ to space were largely consumed by reactions with reduced gases and aqueous species delivered by volcanos and hydrothermal activity. The earliest biosphere depended upon these nonbiological sources of reducing power for the synthesis of organic compounds. Anaerobic microbes dominated the biosphere's contributions to the atmosphere. Some of the key gas-forming microbial processes involved redox reactions, including the synthesis and the degradation of organic compounds. Such processes included methanogenesis, dismutation reactions with carbon and sulfur, anoxygenic photosynthesis, oxidation and reduction of nitrogen species and acetogenesis. These processes created CH₄, larger hydrocarbons, CO, H₂S, (CH₃) ₂S, NH₃, N₂O, methylated halogens and a host of other less abundant, reduced trace gases. The biogenic signature of an atmosphere might consist of a suite of reduced gases whose abundance and diversity are much greater than those in the atmosphere of a lifeless planet.

The second stage of our biosphere's evolution represents the time interval between the evolution of oxygenic photosynthesis and the accumulation of a substantial inventory of oxygen (>0.1 percent by volume) in the global atmosphere and ocean. The advent of oxygenic photosynthesis very likely increased gross global biological productivity by a factor of 100 or 1000. Even so, perhaps hundreds of millions of years were required before the net production of O₂ exceeded the output of reduced volcanic and hydrothermal gases, allowing O₂ to become a dominant atmospheric constituent. During this interval, the atmosphere displayed a very diagnostic biological signature, namely a disequilibrium mixture of reduced gases and O₂.

The third stage encompasses the interval (the past 1.8 to 2 billion years on Earth) where O₂ became a major atmospheric constituent (>0.1 percent by volume). Its production and consumption were dominated by biological processes. Anoxic environments retreated to localities (e.g., fine-grained aquatic sediments, hydrothermal systems, etc.) where they were maintained by local sources of reducing power (e.g., organic decomposition, volcanism). Accordingly, atmospheric concentrations of reduced biogenic gases declined to trace levels. For example, the concentration of CH₄, which is currently the most abundant reduced biogenic gas, is only 1.7 parts per million by volume. Thus, O₂ and O₃ are the most easily detected biogenic gases during this most recent stage.

The above short overview offers several insights. First, life can begin very early in a planet's history. Thus, a planet whose environment is suitable for a biosphere probably developed life relatively quickly. Second, O₂ is not necessarily an atmospheric component shared by all biospheres. Earth required at least 2 to 2.5 billion years of evolution before accumulating an abundant O₂ inventory. Also, even though oxygenic photosynthesis confers substantial advantages for the biota, its development might not necessarily be an inevitable consequence of biological evolution. Third, other reduced gas species are highly diagnostic of life, but their detectability might be constrained by their low concentrations. However, atmospheric budgets of most reduced biogenic gases have not been modeled in O₂-free atmospheres.

Future research should explore the processes responsible for creating unambiguous atmospheric signatures of life. For example, we are still uncertain as to why our own atmospheric O₂ inventory is currently maintained near the 21 percent level. Today, O₂ and its byproducts participate in the destruction of most, if not all, of the reduced biogenic gases. The atmosphere, both of early Earth and, perhaps, of some of the other inhabited planets, had low O₂ contents. Therefore the budgets of reduced gases such as volatile hydrocarbons, sulfides, amines and NO_x should be modelled in low-O₂ atmospheres. Such modelling should explore atmospheric compositions under least the following two scenarios:
1) Oxygenic photosynthesis exists but the volcanic sinks for O₂ are stronger than they are presently on Earth, and 2) Oxygenic photosynthesis does not exist, therefore biological primary productivity is maintained either by abiotic sources of reducing power or by other mechanisms not employed in our own biosphere.

Regarding future research simple ecosystems of bacteria should be manipulated in order to ascertain the various controls upon gas production. For example, the effect of different levels of O₂ and alternative electron acceptors can be evaluated. Such work should be done in collaboration with atmospheric chemists who could evaluate how the gases produced by such ecosystems might be modified subsequent to their emission.

We should define the circumstances under which disequilibrium mixtures of atmospheric species indeed indicate life's presence. Under which circumstances are disequilibrium gas mixtures produced which are both truly diagnostic of life and most easily detected?