There is a grey-green crust on the gravestone, the fencepost, the north face of the boulder. Most people walk past it without registering it as alive, and the ones who do register it file it under moss. It isn’t moss. It isn’t a plant. It isn’t, in the way you mean the word, an it at all. And the question of exactly how many living things you are looking at is genuinely unsettled — not in a hand-wavy way, but in the specific way where working lichenologists have been publishing letters at each other in New Phytologist as recently as 2024.
We’ve spent three issues on fungi as machinery — the kingdom that invented rot, the partnership that built soil, the plumbing under a forest. This one is about a fungus that did something stranger than any of those. It stopped being a thing that lives on and became a thing that lives as. It built a body out of somebody else.
The one-line answer, and why it was heresy
Here is the answer you were taught, from the British Lichen Society: “A lichen is not a single organism; it is a stable symbiotic association between a fungus and algae and/or cyanobacteria” Established. The fungus is the mycobiont. The photosynthesising partner — green alga, cyanobacterium, or both — is the photobiont.
That sentence is now in every textbook. When Simon Schwendener published the equivalent of it, it was an outrage. Lichens had been treated as their own group of plants; in 1869 he wrote that they “do not, from my investigations, constitute a separate, higher division of cryptogams, but merely a subdivision of the great order of Fungi: they are ascomycetes parasitic on algae” Established.
The reaction is worth dwelling on, because it is a very clean specimen of a field defending its furniture. The declaration “sparked a controversy that lasted for almost a quarter of a century.” William Nylander — “foremost lichen taxonomist of his time” — dismissed the whole line of argument in words that reach us at two removes: printed by Annie Lorrain Smith in 1921, from Crombie’s own translation of a later Nylander pronouncement. “All these allegations belong to inept Schwendenerism and scarcely deserve even to be reviewed or castigated so puerile are they—the offspring of inexperience and of a light imagination. No true science there.” The English clergyman-lichenologist James Crombie, reviewing the idea in 1874, complained that it converted lichens “as if by the stroke of a magician’s wand, into a … master fungus, and an imprisoned algal slave,” and argued that it must have been “the very novelty of such a strange theory” — “much more than the apparently plausible arguments by which it was supported” — that recommended it to any botanist of what he called the “new school.” Smith also recorded Crombie dismissing the whole subject as “a Romance of Lichenology, or the unnatural union between a captive Algal damsel and a tyrant Fungal master.”
They were wrong. But notice how they were wrong: not about the microscopy, which everyone could see, but about what the microscopy was allowed to mean. Hold that thought for about a thousand words.
What’s actually in there
Cut a foliose lichen — one of the leafy ones — and you find a layered structure with a proper architecture. There is “a band of compacted hyphae… called a cortex” on top; “below the cortex is a band of photobiont cells and below that is the medulla, an area of loosely arranged hyphae”; below that a second cortex, and from it “root-like bundles of hyphae, called rhizines,” anchoring the whole thing down Established. The algae are not smeared through the fungus at random. They are shelved — held at a specific depth, under a windscreen, at the light level that suits them.
That layout isn’t universal, and the source we’re quoting says so plainly: a crustose lichen has no lower cortex, in a fruticose lichen “it is meaningless to talk of upper and lower sides,” some species have no cortex or no rhizines at all, and in a few genera the photobiont cells are scattered through the thallus rather than banded. We’re describing a common plan, not a law.
The whole structure is called the thallus, and it is “composed of fungal and photobiont cells, so well united as to give the impression that you are looking at just one organism” Established. That impression is the entire problem, and it has been the problem since 1869.
The trade, and who’s getting the better of it
The deal looks straightforward. “The fungi build the structure of the lichen thallus, within which they provide conditions for a long term, stable association with their photobionts”; the fungus takes “simple sugars” and the photobiont gets a habitat, mineral nutrients liberated by fungal digestion, and shelter inside an interior that is “often a place richly infused with complex secondary fungal chemicals found nowhere else in nature” that likely protect “from UV radiation, desiccation, and grazing by herbivores” Established. Where the photobiont is a cyanobacterium, the lichen also gets fixed nitrogen out of thin air Established.
The fungus even does the gardening. It “is able to shift the position of the algal cells over short distances to secure adequate illumination and most efficient photosynthesis” Established. It moves its algae into the light.
You can read that sentence two ways, and lichenology has been reading it both ways for forty years.
Farm, or prison?
In 1981, Vernon Ahmadjian and Jerome Jacobs published resynthesis experiments in Nature and opened the abstract with a line that has aged extremely well: “The nature of the lichen symbiosis is not clear. It is generally thought to be mutualistic but this concept is not supported by experimental evidence.” Their own reading of what they’d grown: “our observations of artificial syntheses of the mycobiont Cladonia cristatella (‘British soldiers’) with different algae suggest that the relationship in this lichen is one of controlled parasitism” Contested. Note what that sentence is and isn’t: a laboratory resynthesis, using algae that weren’t the lichen’s own, not a reading of wild lichens.
The standard textbook lays out both camps without picking one — and hangs the parasitism camp on Ahmadjian’s 1993 book, which we have not opened. “Most general textbooks and many researchers refer to lichens as a classical case of mutualism, where all the partners gain benefits from the association. Alternatively, lichens are regarded as an example of controlled parasitism, because the fungus seems to obtain most of the benefits and the photobiont may grow more slowly in the lichenized state than when free-living (Ahmadjian 1993).” Its verdict on the balance sheet is blunt: “The lichen fungus undoubtedly benefits enormously by obtaining its nutrition from the photobiont, but the photobiont’s gain from the association is less obvious” Contested.
Three concrete things keep the argument alive. First, the accounting: “Up to half of the carbon fixed by algae is immediately converted to fungal sugars which are inaccessible to the alga itself.” Second, the asymmetry of freedom — “though the fungi that form lichens do not occur in nature as independent organisms, a number of the photobionts can be found in free-living forms” Established. One partner can walk away. The other cannot. Third, Ahmadjian never softened: a 2024 review records that he considered lichen algae “a source of food for lichen fungi,” absolutely dependent on the fungal partner, which cultivates them in a way “comparable to heads of lettuce.”
And the mutualists have real answers, which the same sources print. Lichenisation “substantially reduces the light intensity to which the photobiont is exposed” (the textbook credits Ertl 1951), and since high light hurts the photobiont, “lichenization is one mechanism by which photobionts may expand into high light environments. Thus, there may well be benefits to lichenization from the perspective of the photobiont.” The British Lichen Society, laying out both sides, notes that some lichens that form stable associations with their usual algae “form parasitic-type interactions with non-host algae when grown in the lab” — which is a hint that the relationship’s character is not fixed by the fungus alone.
We’re not going to call this one. The field hasn’t.
The experiment that has never quite worked
Here is a fact that ought to be better known, and is a small scandal in the nicest way. Nobody can reliably build a lichen.
A 2025 review in the Annual Review of Microbiology went back through 150 years of attempts and concluded: “despite multiple reports of successful lichen resynthesis, no lichen lab model system exists today” Established. Its verdict on the literature is unsparing but carefully non-accusatory — “the term resynthesis is applied to many types of fungal-photobiont cocultures that do not resemble lichens,” “some of the most lichen-like results, for their part, were obtained from nonaxenic tissue culture,” and of the handful of studies claiming natural-looking lichens from axenic starting cultures, “all appear to have been isolated successes obtained against the background of extensive contamination” — isolated there meaning one-off, not germ-free. Flatly: “Few experiments have ever produced anything resembling a natural lichen.” The reviewers go out of their way not to make this a charge of carelessness: “many researchers took their experiments as far as was possible with available tools,” and judging decades-old axenicity now “is almost impossible and leads to unfruitful speculation.”
What partial success there is comes from making the partners miserable. Early experimenters found that the “early stages of lichen formation begin with nutrient-poor media and conditions that must be unfavorable to the algal symbionts”; a second trick is that “thallus formation is induced by providing wetting-drying cycles” like the ones the lichen meets on a rock — though the same review adds that cyclical drying “was, however, considered unnecessary in some systems.” A 2023 New Phytologist review agrees on the conditions, if not on the scoreboard: “at least a few months up to several years of co-culture of previously isolated primary symbionts are required to achieve re-lichenization into fully formed lichen thalli,” and “cycles of desiccation and rehydration promote in vitro lichenization” Established. Worth flagging that those two reviews are not allies: the 2023 paper treats lab re-lichenization as something that can be achieved, and the 2025 one exists largely to dispute that. Comfort, at any rate, does not seem to be what builds a lichen.
Nor is the 2025 review claiming nothing has ever worked. It flags standing exceptions — Endocarpon pusillum, resynthesised spore-to-spore since 1877 and “replicated at least three times,” and an apparently single Xanthoria parietina thallus that produced apothecia and spores but “lacked the characteristic stratified texture, as well as the characteristic secondary metabolites, of natural X. parietina.” The claim is that no reliable, reproducible model system exists, not that the thing is impossible. And the reviewers are scrupulous about their own position: stringent experiments “might ultimately show that many lichens indeed require only a single fungus and alga to form,” but they “do not consider the experimental evidence to date to be sufficient to reject” the alternative.
And the best story in the whole literature belongs to Eugen Thomas, who in 1939 got one properly differentiated Cladonia chlorophaea thallus, podetia and all, out of a flask left unattended for several months during a stay in Sweden and subsequent military service — then set up eight hundred further cocultures across every pure culture line he had, and never got another differentiated thallus. In the 2025 reviewers’ translation of his German dissertation, he called it a “gift of chance that arose thanks to special, favorable conditions working together in one certain, narrowly circumscribed spot.” Eight hundred tries. One lichen. No idea why.
Which raises the obvious question: if assembling a lichen is that hard, how does the world stay covered in them? Largely by not assembling them. Many lichens reproduce by shipping the partnership out intact — “minute, powdery granules (called soredia), each soredium consisting of a few photobiont cells surrounded by fungal filaments,” or “tiny, simple or branched spiny outgrowths (called isidia), again a mixture of fungal and photobiont cells.” Both are easily broken off and dispersed, and both “contain everything needed to produce new thalli” Established. The crust on the gravestone doesn’t make a baby lichen. It makes a crumb of itself.
Sexual reproduction, meanwhile, belongs to one member only: “Only the fungal partner reproduces sexually,” scattering spores from the little saucers (apothecia) you can see with the naked eye on many species Established. So the lichen’s two modes of continuing are: copy the consortium wholesale, or let the fungus alone have offspring. The partnership never has sex. It is either cloned or re-founded — and given how badly re-founding goes in a laboratory, that is a genuinely strange thing to know about something so common.
The count starts slipping
So much for two. The trouble is that the number has been quietly failing for years, in four different directions at once.
One thallus, several algae. “Photobiont plurality refers to the presence of multiple photobionts within a single lichen thallus. This phenomenon was described at the end of the last century but has received deeper attention in the last decade” Established. As another review puts it: “As opposed to what was traditionally believed, the photobiont should not be limited to a single strain of algae.” The first review is also blunt about how thin the sampling still is: plurality “has been investigated so far in about fifty species of lichen-forming fungi among the approximately 20,000 mycobiont species described,” which “makes interpretation of the data and its generalisation still difficult.” The phenomenon is real. How common it is, nobody yet knows.
Partners get swapped. “Photobiont switching” is the term for a mycobiont relinquishing one photobiont and acquiring another — often along environmental gradients, so that “different photobiont species may be associated with the same mycobiont in lowlands compared to areas a few hundred meters higher” Established. The same lichen “species” is a different partnership at a different altitude.
Some run two power systems. In tripartite lichens, “the fungus is simultaneously associated with both green algae (photobiont) and cyanobacteria (cyanobiont),” the alga distributed through the thallus for carbon and the cyanobacterium “often confined to special structures, called cephalodia,” for nitrogen Established. That’s roughly 3–4% of lichen species running a fungus, an alga and a bacterium in one crust: two domains of life, whose split predates the nucleus itself.
And there are always bacteria. “There is also convincing evidence for a consistent presence of non-photosynthetic bacteria within the thalli of all lichens, although the role of these bacteria is as yet unknown” Established. Present, consistently. Doing what — nobody has established.
An aside on whose name is on the door
One thing that isn’t ambiguous, and that surprises people: when you look up a lichen species, you are looking up the fungus. This isn’t convention or shorthand, it’s codified law. Article F.1.1 of the International Code of Nomenclature for algae, fungi, and plants reads: “For nomenclatural purposes, names given to lichens apply to their fungal component” Established.
Which is a strange thing to sit next to everything above. The formal identity of the organism is assigned to one of its members. Every argument in this issue is, in a sense, an argument with that rule.
And this is not a niche corner of mycology. Lichenisation is one of the great fungal lifestyles: “Fully half of all ascomycetes and one in five of all known fungi form lichens” Established. How many lichen species there are depends on who’s counting — the formal taxonomic tabulation puts “the number of lichenized species… at 19,409,” a government herbarium says “over 20,000 species are known,” and the British Lichen Society says “about 28,000 species worldwide.” Those aren’t measuring quite the same thing, and none of them says so on the page. We’d rather show you three numbers than pick the tidiest one.
Where we get honest
In 2016, a team led by Toby Spribille went looking for why two hair lichens that are genetically inseparable as fungi nonetheless look different — one dark brown, one yellowish with a toxic compound called vulpinic acid. What they found was a second fungus. From the abstract: “many common lichens are composed of the known ascomycete, the photosynthesizing partner, and, unexpectedly, specific basidiomycete yeasts… The structurally important lichen cortex, long treated as a zone of differentiated ascomycete cells, appears to consistently contain two unrelated fungi.” Fluorescent probes lit up “round, ~3- to 4-µm-diameter cells embedded in the peripheral cortex.” Screening “across the seven main radiations of macrolichens in the class Lecanoromycetes,” they found related lineages “associated with 52 lichen genera from six continents” — including 42 of the 56 genera they sampled in a single family, the Parmeliaceae. Genera, note, not species; and macrolichens, not lichens at large. A 2019 follow-up on wolf lichens found a third fungus, a Tremella, in “95% (300 out of 316)” of specimens — meaning most of those lichens carry “a predictable suite of at least three fungal species” Contested.
Two corrections before you repeat that. First: Spribille’s paper never says “third partner.” Its own strongest claim is that the finding “should change expectations about the potential diversity and ubiquity of organisms involved” — the ménage-à-trois framing came from the commentary afterwards — as a 2020 survey put it, the finding “raised speculation of the basidiomycetous yeast being the third mutualistic partner.” Second: the link from the yeast to the vulpinic acid is explicitly unfinished. The authors write that the data “implicate Cyphobasidium in the production of vulpinic acid, either directly or by inducing its synthesis” by the other fungus, and that “confirming a link by using transcriptome or genome data is impossible until the enzymatic synthesis pathway of vulpinic acid is described” Preliminary. That same 2020 survey found that lichen chemotype significantly structured the photobiont community but not the yeast community, and concluded that “direct production of these metabolites by the yeasts seems unlikely” — but “these metabolites” are medullary compounds, from the thallus interior, while vulpinic acid and the yeasts both sit in the cortex, whose chemistry that study could not test at all. Not a head-to-head. The authors also note that their design cannot separate a chemistry effect from a species effect.
And the yeasts have serious critics. A metagenomic sweep of 339 lichen species detected them in “2.7% of all species sampled” and reported that this failure to find them in 97.3% of species “suggests that basidiomycete yeasts are not ubiquitous in lichens.” The 2020 survey found them “much less mycobiont-specific than the photobionts” and suggested “the yeast might not be as intimately associated with the symbiosis as is the photobiont.” Hawksworth and Grube, reviewing the argument in 2020, quote the late Franz Oberwinkler concluding that these yeasts “are not a third component of symbiosis, but rather the vegetative propagules of mycoparasites” — not partners but freeloaders. We have not read Oberwinkler’s 2017 paper ourselves; that sentence reaches us through theirs.
Then, in the same paragraph, those same two authors undercut their own best exhibit: the metagenomic method behind the 2.7% figure “may be much less sensitive than PCR assays with specific primers,” and so it “remained, therefore, unclear, how ubiquitous and specific these yeast asexual stages actually were.” Nor are the two surveys scanning the same shelf. Spribille’s claim was about macrolichens in one fungal class; the metagenomic sweep ran across 57 families in 25 orders, and seven of its nine positives are, on the paper’s own count, foliose or fruticose — macrolichens of exactly the kind Spribille had flagged. A real objection on both sides. Not a resolution on either.
The honest position is the one the 2019 wolf-lichen authors took themselves — and note that they are not simply throwing up their hands: “We still do not know the function of these newly discovered players in lichen symbiosis, but the fact that we were able to achieve in situ rRNA hybridizations of Cyphobasidium and Tremella cells in the Letharia cortex shows that these cells are physiologically active components of that layer.” Present and metabolically alive; job unknown. And then, better still: “it may be more honest to concede we do not know the function of a fungus, even if this temporarily impinges on long-standing assumptions regarding the minimum components of a lichen.” We’ll take that sentence over a tidy headline any day.
Which leaves the fight over the word itself — and this one isn’t an experiment anybody can run. In 2020 David Hawksworth and Martin Grube proposed: “A lichen is a self-sustaining ecosystem formed by the interaction of an exhabitant fungus and an extracellular arrangement of one or more photosynthetic partners and an indeterminate number of other microscopic organisms” Philosophy. In 2024 William Sanders wrote back that this “elevates microbiome inhabitants to a defining status equivalent to that of the mycobiont and photobiont(s), even though their roles and significance are still poorly understood and their critical importance to the symbiosis far from proven”; that “although microbiome research has changed how we think about plants and animals, it has not resulted in any effort to redefine them”; that “ecosystem” is technically the wrong word, since the term as coined includes the physical substratum and “for most biologists, a lichen removed from its substratum is still a lichen”; and — this last aimed not at Hawksworth and Grube but at a different widening proposed by Lücking and colleagues in 2021, a paper we have not opened and know only as Sanders describes it — that building “plausible but entirely hypothetical cases” into the definition of a lichen carries “a significant risk of circular reasoning.” His bottom line: “Subtract the mycobiont or photobiont, by contrast, and lichen identity is clearly lost. They are the defining elements, at least for now.” Hawksworth and Grube replied that lichens deserve the word because they are “self-contained, functionally independent operating units,” unlike a mycorrhiza, which “refers to just part of the plant and part of the fungus.”
We’ve stamped that exchange Philosophy rather than Contested on purpose, and it was a near thing, so here is the reasoning. What’s in the thallus is empirical, and more data will move it. Whether that content warrants rewriting the word is a judgement about where to draw a line: neither man disputes the evidence the other cites — Sanders says as much, that it “comes as a surprise that they feel the definition of a lichen needs to be amended.” What keeps it from being pure philosophy is that Sanders has published the terms of his own surrender: “If it can be demonstrated that additional microorganisms are essential to the transformative development of the lichen thallus, then their inclusion in the definition might be justified.” So: definitional at the surface, empirically gated underneath. We’ve stamped the surface, and we’re telling you about the gate.
Which is the thing that has us thinking. Schwendener’s critics weren’t denying what was under the lens; they were refusing a new place to draw the line around an individual. A hundred and fifty years later the lens has got much better and the line is wobbling again. Where does one organism end? We’re not answering that here — it’s a whole series in this plan, and it deserves the room. But if you want the shortest honest statement of what a lichen is, it might be this: a lichen is what happens when the question stops having a clean answer.