The earliest photosynthetic organisms were almost certainly bacteria that harvested light without producing oxygen, using electron donors like hydrogen, iron, or sulfide instead of water. Geochemical signatures consistent with biological carbon fixation appear in rocks as old as 3.5 billion years, and the weight of evidence points to anoxygenic phototrophs as the pioneers. The more familiar oxygen-producing version of photosynthesis, performed by cyanobacteria, likely came later, though exactly how much later remains one of the most contested questions in evolutionary biology.
What the Carbon Isotope Record Shows
The oldest chemical fingerprints of photosynthesis come not from fossils you can see, but from the ratio of carbon isotopes locked in ancient rocks. Photosynthetic enzymes preferentially grab the lighter form of carbon, so organic matter produced by photosynthesis ends up enriched in that lighter isotope compared to the carbonate minerals deposited from seawater at the same time. This isotopic gap shows up consistently in rocks dating to roughly 3.2 to 3.5 billion years ago, matching the fractionation pattern expected from the enzyme that fixes carbon in cyanobacteria and plants today.1PubMed Central. The paleobiological record of photosynthesis
Even older carbon isotope signals appear in graphite from the roughly 3.8-billion-year-old Isua formation in Greenland. But researchers have attributed those signatures to hydrogen-driven photosynthesis rather than the oxygen-producing kind, reasoning that there would have been no evolutionary pressure to split water when hydrogen gas was freely available as an electron source.2Photosynthesis Research. Photosynthesis in the Archean era If that interpretation holds, it pushes back the origin of at least simple light-driven metabolism to near the end of the Late Heavy Bombardment, when Earth was barely habitable by any modern standard.
Stromatolites and the Earliest Visible Life
Carbon isotopes are invisible to the naked eye, but some of the earliest evidence for photosynthetic communities comes in a form you can walk across. The Strelley Pool formation in Western Australia, roughly 3.43 billion years old, preserves kilometer-scale remnants of a stromatolitic carbonate platform.3Precambrian Research. 3.43 billion-year-old stromatolite reef from the Pilbara Craton of Western Australia: Ecosystem-scale insights to early life on Earth Stromatolites are layered structures built up by microbial mats, typically photosynthetic ones, trapping and binding sediment as they grow toward the light.
At Strelley Pool, researchers identified seven distinct stromatolite shapes across different parts of what was once a shallow tidal carbonate platform. The diversity of forms and their environmental associations mirror patterns seen in younger, unambiguously biological stromatolites, which the researchers argued rules out non-biological explanations for the structures.4Nature. Stromatolite reef from the Early Archaean era of Australia Closer inspection of well-preserved outcrops revealed microscale textures, organic layers, and fabric patterns that link the different stromatolite shapes to changes in sedimentation and microbial mat development.5PubMed Central. Controls on development and diversity of Early Archean stromatolites
Whether the microbes building those 3.43-billion-year-old reefs were producing oxygen or running on some other electron donor is an open question. The existence of complex reef ecosystems at that time at least tells us that photosynthetic communities were already ecologically dominant in shallow marine environments.
Why Anoxygenic Phototrophs Likely Came First
The conventional view for decades has been that anoxygenic photosynthesis preceded the oxygen-producing kind. Organisms like green sulfur bacteria, purple bacteria, and heliobacteria use a single photosystem and draw electrons from compounds like hydrogen sulfide, ferrous iron, or even arsenic, rather than water. Because splitting water is thermodynamically harder and requires more complex molecular machinery, the assumption has long been that simpler forms of photosynthesis arose first and the water-splitting apparatus evolved later.
That logic is persuasive, but not everyone is convinced the case is closed. A provocative review has argued that neither the molecular evolutionary data nor the geochemical record has conclusively proven anoxygenic photosynthesis preceded oxygenic photosynthesis, and that much of the confidence in the standard timeline reflects an interpretive bias carried forward from early speculation.6PubMed Central. Thinking twice about the evolution of photosynthesis The point is not that the conventional sequence is wrong, but that it may be less iron-clad than textbooks suggest. Researchers have been debating for years whether the two photosystems in cyanobacteria arose from an internal gene duplication before photosynthesis spread across the bacterial tree, or were acquired from different anoxygenic lineages through horizontal gene transfer relatively late in cyanobacterial history.
Banded Iron Formations and Iron-Eating Phototrophs
Some of the strongest circumstantial evidence for ancient anoxygenic photosynthesis comes from banded iron formations, or BIFs, those striking layered rocks of alternating iron-rich and silica-rich bands found in Archean and early Proterozoic sediments worldwide. Something was oxidizing the dissolved ferrous iron in the ancient ocean to produce the insoluble iron minerals in BIFs, and the two main candidates are oxygenic photosynthesis (where the oxygen produced by cyanobacteria rusted the iron chemically) and photoferrotrophy, an anoxygenic process in which bacteria use ferrous iron directly as their electron donor.7PubMed Central. Photoferrotrophy, deposition of banded iron formations, and methane production in Archean oceans
Modern analogue studies strengthen the case for photoferrotrophy. In a stratified lake with iron-rich, oxygen-free deep water, researchers found thriving communities of photoferrotrophic green sulfur bacteria at the boundary layer where light penetrated but oxygen did not. The team proposed that similar microbial communities populated the boundaries of ancient iron-rich oceans, driving BIF deposition and fueling early marine productivity.8PubMed Central. Photoferrotrophs thrive in an Archean Ocean analogue Geochemical work on specific BIF deposits has reinforced the idea that anaerobic microbial iron oxidation played a major role in BIF formation before the rise of atmospheric oxygen.9Evolving Earth. Anaerobic photoferrotrophy and hydrothermal influence in the Mesoarchean Girar banded iron formation: insights from micropaleontology and geochemistry
If photoferrotrophs were indeed responsible for much of the iron oxidation in the early ocean, they were not just surviving but reshaping global geochemistry billions of years before oxygen-producing photosynthesis took over.
Heliobacteria and the Simplest Reaction Center
Among living phototrophs, heliobacteria occupy a special position in the search for the ancestral form of photosynthesis. They possess what is arguably the simplest known pigment-protein complex capable of converting light into chemical energy.10PubMed. Thermodynamics of the Electron Acceptors in Heliobacterium modesticaldum: An Exemplar of an Early Homodimeric Type I Photosynthetic Reaction Center Their reaction center is a homodimer, meaning both halves are identical copies of the same protein, a symmetry that many researchers consider a retained ancestral feature rather than a later simplification.
The crystal structure of the heliobacterial reaction center, solved at high resolution, reveals perfect two-fold symmetry. It coordinates dozens of pigment molecules that capture and funnel light energy to an electron transfer chain at its center, which includes an iron-sulfur cluster. Unlike other reaction centers, it lacks a bound quinone molecule.11PubMed. Structure of a symmetric photosynthetic reaction center-photosystem Researchers have described this structure as preserving characteristics of the ancestral reaction center, making heliobacteria a living window into how photosynthesis may have looked in its earliest incarnation.
Structural analysis has also confirmed that both major types of photosynthetic reaction centers, despite looking quite different on the surface, share a common evolutionary origin traceable to a single ancestral protein.12PubMed. A cytochrome b origin of photosynthetic reaction centers: an evolutionary link between respiration and photosynthesis The divergence between these two types is a key chapter in the story, because cyanobacteria are the only organisms that wired both types together, enabling the water-splitting chemistry that produces oxygen.
When Did Oxygen-Producing Photosynthesis Appear
The most visible consequence of oxygenic photosynthesis is the Great Oxidation Event, the period roughly 2.3 to 2.4 billion years ago when oxygen first accumulated in the atmosphere to detectable levels. But the organisms producing that oxygen almost certainly evolved well before the atmosphere reflected their presence. Sulfur isotope data suggest that oxygenic photosynthesis was already a major contributor to biological productivity by about 2.7 billion years ago, several hundred million years before oxygen levels permanently rose.13Earth and Planetary Science Letters. Atmospheric sulfur rearrangement 2.7 billion years ago: Evidence for oxygenic photosynthesis
Molecular clock analyses place the origin of cyanobacteria as a group even earlier. One study dated the last common ancestor of crown-group cyanobacteria to between about 3.3 and 2.7 billion years ago, depending on which fossil calibrations and evolutionary models were applied.14PubMed Central. The Archean origin of oxygenic photosynthesis and extant cyanobacterial lineages That broad range illustrates the difficulty of pinning down exact dates from genetic data alone, but the estimates at least overlap with the oldest stromatolite and carbon isotope evidence.
Even after cyanobacteria began producing oxygen, it took a long time for that oxygen to stick around. Atmospheric oxygen levels appear to have bounced above and below a critical threshold for roughly 200 million years after about 2.32 billion years ago, with permanent oxygenation arriving only around 2.22 billion years ago, coinciding with a major carbon isotope excursion.15Nature. A 200-million-year delay in permanent atmospheric oxygenation The gap between the biological innovation and its atmospheric signature reflects how effectively oxygen was consumed by reactions with iron, sulfur, and methane in the early ocean and atmosphere.
Non-Photosynthetic Relatives of Cyanobacteria
A surprising twist in the story came from genomics. Researchers sequencing microbial genomes from the human gut and other oxygen-poor environments discovered a class of bacteria, called Melainabacteria, that are closely related to cyanobacteria but completely lack photosynthetic machinery.16PubMed Central. Non-Photosynthetic Melainabacteria (Cyanobacteria) in Human Gut: Characteristics and Association with Health An even more basal class, Sericytochromatia, was subsequently identified, and it too is non-photosynthetic. Because both lineages that branch off before the photosynthetic cyanobacteria lack any photosynthetic genes, the implication is that phototrophy was not an ancestral feature of the broader cyanobacterial lineage. Instead, the photosynthetic branch acquired its light-harvesting genes relatively late in cyanobacterial evolution.17PubMed. On the origins of oxygenic photosynthesis and aerobic respiration in Cyanobacteria
This finding complicates the narrative. If even cyanobacteria did not start out as phototrophs, then both the machinery for anoxygenic photosynthesis and the upgrade to oxygenic photosynthesis may have been acquired events rather than inheritance from some single ancient photosynthetic ancestor. It also raises the question of where the photosynthesis genes originally resided before they made their way into the cyanobacterial lineage.
Horizontal Gene Transfer Scrambles the Timeline
One reason the evolutionary history of photosynthesis is so hard to untangle is that the genes for it have moved sideways between unrelated lineages repeatedly. The photosynthesis gene cluster has been documented transferring between different classes of purple bacteria, with subsequent rearrangement of the genes within the receiving organism.18PubMed. Horizontal transfer of the photosynthesis gene cluster and operon rearrangement in purple bacteria Even more striking, the discovery of photosynthetic capability in the phylum Gemmatimonadetes, a group not previously known to contain phototrophs, was attributed to an ancient horizontal transfer from purple bacteria. That made Gemmatimonadetes the seventh bacterial phylum known to contain species with chlorophyll-based photosynthesis.19PubMed Central. Functional type 2 photosynthetic reaction centers found in the rare bacterial phylum Gemmatimonadetes
The consequence of widespread horizontal transfer is that a family tree of photosynthesis genes does not necessarily match the family tree of the organisms that carry them. You cannot simply look at which bacteria are most distantly related and conclude that their shared ancestor invented photosynthesis. The genes may have arrived by a completely separate route. Molecular phylogenetic analysis of pigment synthesis pathways has concluded that bacteriochlorophyll biosynthesis evolved before chlorophyll biosynthesis, arguing against the older idea that simpler pigments came first and were later modified into more complex ones.20PubMed. Molecular evidence for the early evolution of photosynthesis But horizontal gene transfer means even that conclusion has to be held with some caution.
How the Water-Splitting Machinery May Have Originated
The transition from anoxygenic to oxygenic photosynthesis required the invention of one of the most remarkable catalytic centers in biology: the manganese-calcium cluster that rips electrons from water molecules. This cluster, sitting at the heart of Photosystem II, contains four manganese atoms and one calcium atom held together in a specific geometric arrangement. How did something so precise evolve?
One hypothesis proposes that the proto-enzyme derived from naturally occurring manganese oxide minerals that precipitated in the early ocean. Several classes of manganese oxide minerals found in nature contain structural motifs similar to the biological cluster.21PubMed Central. A possible evolutionary origin for the Mn4 cluster of the photosynthetic water oxidation complex from natural MnO2 precipitates in the early ocean Building on this idea, a separate study showed that amorphous calcium-manganese oxides, which could have formed in Archean oceans as pH shifted from acidic to alkaline conditions, are effective catalysts for water oxidation in the lab. The bond types and lengths in these simple mineral compounds are directly comparable to those in the biological oxygen-evolving complex, suggesting that early cyanobacteria may have essentially co-opted a mineral catalyst and refined it with amino acid scaffolding.22PubMed. Amorphous manganese-calcium oxides as a possible evolutionary origin for the CaMn₄ cluster in photosystem II
Beyond Chlorophyll: Rhodopsin-Based Light Harvesting
Chlorophyll-based photosynthesis is not the only way organisms have learned to use light. Microbial rhodopsins, simple membrane proteins related to the light-sensing pigments in animal eyes, are widespread among bacteria and archaea and function as light-driven proton pumps. They do not fix carbon the way chlorophyll-based photosynthesis does, but they convert light energy into a proton gradient that cells can use to make energy currency.
Reconstructions of ancestral rhodopsin sequences suggest that the earliest versions absorbed green light, with predicted absorption peaks between 537 and 543 nanometers, and functioned as proton pumps. Green-light absorption would have allowed their hosts to live at depths where damaging ultraviolet light was filtered out but visible light still penetrated.23Molecular Biology and Evolution. Earliest Photic Zone Niches Probed by Ancestral Microbial Rhodopsins Whether rhodopsin-based phototrophy is older than chlorophyll-based photosynthesis, or whether both emerged around the same time and simply occupied different ecological niches, remains an open question. What is clear is that the exploitation of light energy is not a single innovation but has been invented independently more than once.
Exotic Electron Donors on the Early Earth
The standard textbook list of electron donors for anoxygenic photosynthesis includes hydrogen, hydrogen sulfide, and ferrous iron. But the early Earth offered some less familiar options. In the hot springs feeding into Mono Lake, California, researchers discovered microbial biofilms carrying out light-dependent oxidation of arsenite to arsenate under completely oxygen-free conditions. A pure culture of the responsible photosynthetic bacterium grew using arsenite as its sole electron donor.24PubMed. Arsenic(III) fuels anoxygenic photosynthesis in hot spring biofilms from Mono Lake, California Phylogenetic analysis indicated that microbial arsenic metabolism is ancient and probably extends back to the earliest Earth.
Findings like this broaden the picture of early photosynthetic ecology considerably. The earliest phototrophs were not limited to one or two chemical fuels. They exploited whatever reductants were available in their local environment, and the diversity of electron donors used by modern anoxygenic phototrophs likely reflects a long history of metabolic experimentation stretching back billions of years.
How Photosynthesis Reached Eukaryotes
For roughly the first two billion years of its history, photosynthesis was an exclusively bacterial affair. Eukaryotes, the larger and more complex cells that include everything from algae to oak trees, did not invent photosynthesis on their own. They acquired it by engulfing a cyanobacterium and retaining it as an internal symbiont. This event, called primary endosymbiosis, occurred in the common ancestor of the group that includes green algae, red algae, and an obscure group called glaucophytes, and it appears to have happened only once.25PubMed Central. Primary endosymbiosis and the evolution of light and oxygen sensing in photosynthetic eukaryotes
That single event then branched out in unexpected ways. Other eukaryotes that missed the primary endosymbiosis got a second chance by engulfing algae that already had plastids, in what is called secondary endosymbiosis. Green algal plastids were picked up by euglenids and chlorarachniophytes. Red algal plastids were apparently taken up once, giving rise to a hugely diverse group that includes diatoms, brown algae, and many dinoflagellates. Some dinoflagellates went further still, replacing their inherited plastids through tertiary endosymbiosis with other algae.26PubMed Central. The endosymbiotic origin, diversification and fate of plastids The result is a tangled web of plastid origins that makes the photosynthetic family tree of eukaryotes far messier than the bacterial one.
The practical consequence is that the green color of a forest and the brown color of kelp trace back, through very different routes, to that same ancient cyanobacterial ancestor. Every oxygen-producing photosynthetic organism on Earth today, from a diatom to a redwood, owes its light-harvesting ability to a bacterial lineage that figured out how to split water sometime in the Archean eon.