Microbial organisms dominate life on Earth, but tracing their early history and evolution is difficult because they rarely fossilize. Determining when exactly a particular group of microbes first appeared is especially hard. However, ancient sediments and rocks hold chemical clues of available nutrients that could support the growth of bacteria.

A key turning point was when oxygen accumulated in the atmosphere around 2.3 billion years ago. Scientists have used this oxygen surge and how microbes adapted to it to map out bacterial evolution.

In a study published in Science, researchers from the Model-Based Evolutionary Genomics Unit at the Okinawa Institute of Science and Technology (OIST) and their international collaborators have constructed a detailed timeline for bacterial evolution and oxygen adaptation.

Their findings suggest some bacteria could use trace oxygen long before evolving the ability to produce it through photosynthesis.

The researchers focused on how microorganisms responded to the Great Oxygenation Event (GOE) some 2.3 billion years ago. This event, triggered in large part by the development of oxygenic (oxygen-generating) photosynthesis in cyanobacteria and carbon deposition, fundamentally changed Earth's atmosphere from one mostly devoid of oxygen to one where oxygen became relatively abundant, as it is today.

Until now, establishing accurate timescales for how bacteria evolved before, during, and after this pivotal transition has been difficult due to incomplete fossil evidence and the challenge of determining the maximum possible ages for microbial groups—given that the only reliable maximum limit for the vast majority of lineages is the moon-forming impact 4.5 billion years ago, which likely sterilized the planet.

The researchers addressed these gaps by concurrently analyzing geological and genomic records. Their key innovation was to use the GOE itself as a time boundary, assuming that most aerobic (oxygen-using) branches of bacteria are unlikely to be older than this event—unless fossil or genetic signals strongly suggest an earlier origin. Using Bayesian statistics, they created a model that can override this assumption when data supports it.

This approach, however, requires making predictions about which lineages were aerobic in the deep past. The team used probabilistic methods to infer which genes ancient genomes contained, and then machine-learning to predict whether they used oxygen.

To best utilize the fossil record, they leveraged fossils of eukaryotes, whose mitochondria evolved from Alphaproteobacteria, and chloroplasts evolved from cyanobacteria to better estimate how and when aerobic bacteria evolved.

Their results indicate that at least three lineages had aerobic lifestyles before the GOE—the earliest nearly 900 million years before—suggesting that a capacity for using oxygen evolved well before its widespread accumulation in the atmosphere.

Intriguingly, these findings point to the possibility that aerobic metabolism may have occurred long before the evolution of oxygenic photosynthesis.

Evidence suggests that the earliest aerobic transition occurred in an ancestor of photosynthetic cyanobacteria, indicating that the ability to utilize trace amounts of oxygen may have allowed the development of genes central to oxygenic photosynthesis.

The study estimates that the last common ancestor of all modern bacteria lived sometime between 4.4 and 3.9 billion years ago, in the Hadean or earliest Archean era. The ancestors of major bacterial phyla are placed in the Archean and Proterozoic eras (2.5–1.8 billion years ago), while many families date back to 0.6–0.75 billion years ago, overlapping with the era when land plants and animal phyla originated.

Notably, once atmospheric oxygen levels rose during the GOE, aerobic lineages diversified more rapidly than their anaerobic counterparts, indicating that oxygen availability played a substantial role in shaping bacterial evolution.

"This combined approach of using genomic data, fossils, and Earth's geochemical history brings new clarity to evolutionary timelines, especially for microbial groups that don't have a fossil record," Prof. Gergely Szöllősi, leader of the Model-based Evolutionary Genomics Unit, highlighted.

"Our work also shows that modeling microbial traits from their genomes using machine learning works well for studying the spread of aerobic metabolisms and might also be a useful approach for exploring how other traits emerged and interacted with the planet's shifting environment across geological time," Dr. Tom Williams, a researcher from the University of Bristol's School of Biological Sciences, explained.