Fungi

A mycelium grows by extending the tips of its filaments, and it has no centre — no brain, no trunk, nothing that could be called the main body. When a cord-forming fungus in soil finds a new piece of wood, it thickens the connections leading to it and withdraws material from the directions that found nothing. The network continually rebuilds itself according to what it has encountered.

It is tempting to call this intelligence and it is the wrong word, for a reason worth being precise about. There is no representation of the problem and nothing making a decision. There is local reinforcement of what worked and local withdrawal from what did not, and efficient allocation falls out with no allocator. That is a more interesting fact than intelligence would be, because it is a mechanism rather than a comparison.

A note on words, because two of them cause trouble. "Toadstool" has no biological meaning at all: there is no group of fungi it picks out, no line separating a mushroom from a toadstool, and no way to use it as a safety guide. It is a folk term that in practice means "a mushroom I do not trust", and the trust is not based on anything. And a fungus is not a plant — it has no chlorophyll, makes no food from sunlight, and feeds by digesting things outside itself. A mushroom is not a fruit either, since fruit is a structure of flowering plants; it is a fruiting body, which is a different word doing a similar job.

The comparison that gives this force: around 400,000 plant species have been described, and around 120,000 fungal ones — yet the true fungal total is probably several times the plant total. The kingdom people know least about is very likely the larger one.

A range spanning nearly a factor of two is not a failure of the estimate; it is an honest report of what the methods support. Ratio-based approaches, environmental sequencing and description rates each give a different answer, and the range is where they overlap.

Lichens were the founding example of symbiosis, described as one fungus plus one alga since the 1860s, and taught that way ever since. In 2016 a puzzle forced the question open: two North American lichens that look different, smell different and differ chemically turned out to be identical by every fungal and algal marker.

Sequencing everything present found a basidiomycete yeast sitting in the outer layer of both, in abundances that tracked the chemical difference — and then in macrolichens on six continents. That two of the most-handled organisms in North American forests had a third partner nobody had noticed is a useful correction about what "settled" means in biology.

Lignin is the compound that makes wood stiff and is deliberately difficult to break down. Almost nothing degrades it efficiently except a group of fungi that evolved enzymes for the job, and that fact has a geological consequence: for as long as nothing could digest lignin, dead trees accumulated instead of decaying.

Reconstructing when those enzymes appeared across thirty-one fungal genomes puts their expansion around the end of the Carboniferous — the point at which the rate of coal formation fell sharply. The coincidence is striking. It is also contested, because the same period saw changes in climate and in the basin conditions coal formation requires, and molecular dating carries wide intervals. NatureHQ records it as a compelling hypothesis rather than a settled cause.

A mycelium spreads perfectly well without ever producing a mushroom, and most of the time it does not produce one. The mushroom is built for a single job: to lift spores off the ground and let go of them somewhere the air is moving. Everything about its shape follows from that. The stem is a delivery system for height. The cap is an umbrella keeping the spore-bearing surface dry. The gills are a way of packing an enormous amount of surface into a small object — a field mushroom carries several hundred square centimetres of spore-producing tissue in something that fits in a hand.

It is also expensive. A mushroom is assembled in a day or two from water and stored resources, and it is thrown away just as fast. That cost is why fruiting is triggered rather than continuous: it takes a combination of a mature enough network, sufficient reserves, the right temperature, and — for most species — rain. This is why mushrooms appear in flushes after wet weather rather than gradually. The fungus was there all along; the conditions for spending the reserves were not.

Not every fungus does it this way. Puffballs hold their spores inside until something strikes them and puffs them out. Stinkhorns wrap theirs in foul-smelling slime and let flies carry them off. Truffles fruit underground and depend entirely on being smelled out, dug up and eaten, which is why they smell the way they do. Each of these is a different answer to the same problem, and each abandons the gill-and-catapult mechanism entirely.

The problem the mushroom has to solve is not distance. It is the layer of completely still air that clings to any surface. A spore released into that layer would simply sit there. So the launch has to be violent enough to cross a boundary a fraction of a millimetre thick, and no more — a spore thrown any harder would hit the opposite gill.

The mechanism is a surface-tension catapult and it uses no muscle, no pressure and no moving parts. The spore sweats a droplet at the point where it is attached. A film of water spreads over the spore body. When the two touch, they merge in microseconds, the centre of mass jumps, and the spore is flung off — at accelerations that have been measured in the tens of thousands of g. It travels roughly a tenth of a millimetre, falls down between the gills, and from there the wind does the rest.

This explains two things a mushroom does that otherwise look fussy. Gills must be vertical, or the spore will not fall clear once launched — which is why a mushroom picked and laid on its side will bend its stem overnight to point the cap back down. And the whole system needs humid air to condense the droplet, which is why spore release peaks in damp conditions and stops in dry ones.

A spore lands and germinates into a mycelium with one nucleus per cell — haploid, a single set of chromosomes, and unable to fruit on its own. It grows until it meets another mycelium. If the two differ at both mating-type loci they fuse, and what happens next has no animal equivalent: the nuclei do not merge. They pair up and travel together, two distinct genomes side by side in every cell, dividing in step. This is the dikaryon, and it is the state a mushroom-forming fungus spends most of its life in.

So the answer to whether a mycelium is haploid or diploid is usually neither. It is dikaryotic — two haploid nuclei per cell, not one diploid nucleus. The genomes only combine at the very end, in the cells lining the gills, and are immediately separated again by meiosis into the four spores each of those cells releases. The diploid phase of a mushroom’s life can be measured in hours, and it happens inside a structure that is about to be thrown away.

The mating-type system is what makes this workable. Two unlinked loci, each with many alleles, mean a spore is compatible with the large majority of unrelated mycelia it might meet, while still being incompatible with its own siblings — outcrossing without needing two sexes. In *Schizophyllum commune* the number of mutually compatible types runs into the thousands. That figure is the source of the "fungi have thousands of sexes" headline, which is fun and is not what the system is.

Plenty of fungi never do any of this. Many reproduce clonally, dispersing spores produced by ordinary cell division, and some have no known sexual stage at all — a group that for most of the twentieth century was classified separately for exactly that reason, until genetics showed they belonged in the same lineages as everything else.

The comparison to seeds is the most common way of explaining spores and it quietly gets the economics backwards. A seed is a packed lunch with an embryo in it: fertilisation has already happened, the young plant is built, and the reserves to start it are included. That is expensive, so a plant makes hundreds or thousands. A spore is one cell with almost nothing in it, so a mushroom makes billions. Same job, opposite bet.

The word is also doing three different jobs across biology, and conflating them is a genuine error rather than a pedantic one:

A bacterial endospore is the one that is genuinely not reproduction at all. The bacterium builds an armoured version of itself inside its own cell wall and shuts down; when conditions improve, one cell wakes up. Nothing has been multiplied. Anthrax and botulism persist in soil for decades on exactly this mechanism, which is why the distinction matters outside a classroom.

Once the branching order is right, several things that looked like coincidences stop being coincidences. Fungi store energy as glycogen, the same molecule an animal stores it as, rather than as starch. They build their walls out of chitin, which is what an insect builds its skeleton from. And they feed by digesting other organisms — externally, by releasing enzymes and absorbing what comes back, but by digestion, which is what an animal does inside a gut. A fungus is not a stationary plant. It is a stationary organism that eats.

A mycelium has a real coordination problem: it may cover square metres, it has no nerves and no brain, and its far edges must behave as parts of one organism. It solves this — cords thicken towards food, exhausted regions are abandoned and their material recovered, whole networks fruit at once — and chemistry, hyphal fusion and electrical activity are all involved. The claim that the electrical activity constitutes a language with a vocabulary of fifty words is a separate matter, and it does not survive contact with the method that produced it.

A fungus cannot move, in the sense of picking itself up. It moves by growing — hyphae extend only at their tips, so the network advances at the front while the older parts behind may be dying back. Over a season this produces real travel across a forest floor, and a fairy ring is the visible record of it: a mycelium expanding outward from a starting point, fruiting at the growing edge, with the exhausted centre left behind. Some rings have been measured expanding for centuries.

There is one genuine exception at the microscopic scale. Zoospores — the swimming spores of chytrid fungi — have a flagellum and move under their own power through water. They are a reminder that the fungal ancestor was aquatic, and the same trait in a related lineage is a reason fungi are grouped with animals rather than plants.

Lifespan is a harder question than it looks, because it depends on what you decide the organism is. A mushroom lasts days. A yeast cell divides a few dozen times and stops. But a mycelium has no fixed body to wear out — it grows at the edges and abandons the middle, so there is no whole to age. The Michigan *Armillaria* genet is estimated at over 1,500 years old, and that estimate is arrived at by measuring genetic identity across a wood and dividing by a growth rate, not by finding anything old.

Almost every major correction in this history has the same shape. A group was classified by what it looked like, the classification held for decades because nobody had a way to look past appearance, and then a method arrived — a microscope, a genome, a sequencer pointed at everything present rather than at what was expected — and the group turned out to be something else. It is worth reading with the present tense in mind, because the current picture rests on the same kind of evidence as the pictures it replaced.

1. 1729

   First observation

   Micheli shows fungi grow from spores

   Pier Antonio Micheli sows spores on cut melon and gets the same fungi back, against a prevailing view that moulds arose spontaneously from decaying matter. The result was largely set aside for over a century.

2. 1867

   Reinterpretation

   Schwendener proposes that a lichen is two organisms

   Simon Schwendener argues that a lichen is a fungus cultivating an alga rather than a single organism. Leading lichenologists rejected it for roughly two decades before the evidence became difficult to argue with.

3. 1885

   First observation

   Frank names the mycorrhiza

   Albert Bernhard Frank describes the fungal sheath on tree roots and proposes it is a mutual arrangement rather than an infection. He coins "mycorrhiza" for it.

4. 1969

   Reinterpretation

   Fungi are given their own kingdom

   Whittaker’s five-kingdom scheme removes fungi from the plants on the grounds that they absorb their food rather than making it. The separation is right; the placement beside plants is not.

   Changes how the 1867 result reads

   Schwendener’s lichen and Frank’s mycorrhiza had both shown fungi living inside relationships that plant taxonomy had no room for. A separate kingdom was the first structural admission of it.

5. 1992

   Modern discovery

   A single fungus is mapped across fifteen hectares

   Genetic markers show that one Armillaria gallica individual in a Michigan forest occupies about fifteen hectares — evidence that the organism is the network, not the mushrooms above it.

   The fungus Armillaria bulbosa is among the largest and oldest living organisms

6. 1993

   Reinterpretation

   Molecular phylogeny puts fungi nearer animals than plants

   Sequence comparisons place fungi and animals as sister groups. A kingdom filed for three centuries alongside plants sits on the animal side of the tree.

   Changes how the 1969 result reads

   The 1969 scheme separated fungi from plants for a defensible reason — how they feed — but kept them adjacent. Molecular data indicated the split is far deeper than the scheme implied.

   Monophyletic origins of the metazoa: an evolutionary link with fungi

7. 1997

   Landmark experiment

   Carbon is traced moving between trees through fungi

   Isotope labelling shows net carbon transfer between paper birch and Douglas fir seedlings sharing ectomycorrhizal fungi. The measurement is solid; the popular reading of it is where the trouble starts.

   Net transfer of carbon between ectomycorrhizal tree species in the field

8. 2003

   Replication

   A larger genet is mapped, and the definition starts to strain

   Systematic sampling across a mixed-conifer forest in Oregon maps one Armillaria ostoyae individual across roughly 965 hectares — nearly ten square kilometres — with an age extrapolated at somewhere between 1,900 and 8,650 years.

   Changes how the 1992 result reads

   Sixty times the area, and the question changes shape with the scale. At fifteen hectares "one organism" is a striking finding; at ten square kilometres it becomes a question about the word, because genetic identity across an area does not establish that the mycelium is physically continuous through all of it.

   Coarse-scale population structure of pathogenic Armillaria species in a mixed-conifer forest in the Blue Mountains of northeast Oregon

9. 2016

   Modern discovery

   The lichen has a third partner

   Sequencing everything present in two chemically distinct but genetically identical lichens finds a basidiomycete yeast in the outer layer, and then in macrolichens on six continents.

   Changes how the 1867 result reads

   The two-partner account was not overturned. It was shown to be incomplete in the best-studied symbiosis in biology, after 150 years in which the question was treated as closed.

   Basidiomycete yeasts in the cortex of ascomycete macrolichens

10. 2016

    Replication

    The exchange is measured in mature forest

    Isotope work in tall trees quantifies bidirectional carbon movement between neighbouring species through shared fungi, and describes it as a fungal-mediated exchange rather than trees feeding each other.

    Changes how the 1997 result reads

    Replication supported the movement of carbon. It did not support the intentional reading — trees deciding to help their neighbours — which is the part that travelled furthest.

    Belowground carbon trade among tall trees in a temperate forest

11. 2017

    Modern discovery

    The size of the kingdom is re-estimated upward

    A synthesis of sampling and sequencing data puts plausible fungal diversity at 2.2 to 3.8 million species against roughly 120,000 described, which would leave most fungi unnamed.

    Fungal diversity revisited: 2.2 to 3.8 million species

12. 2023

    Challenge

    The network literature is audited for citation drift

    A review of how common mycorrhizal network claims are cited finds the strength of the evidence tends to grow as it is passed along, and that several widely repeated statements are not supported by the studies given for them.

    Changes how the 1997 result reads

    The challenge is aimed at the citation record rather than at the original measurements. That distinction matters: the 1997 result stands, and what has been built on top of it is what is under review.

• Did fungal lignin decomposition end the Carboniferous coal deposits, or did geology?

  Why it matters: It is one of the few proposed cases of a single evolutionary innovation changing the planet’s geological record, and it is genuinely unresolved.

• What does the lichen yeast actually do?

  Why it matters: Its presence is established and its function inferred, so the three-partner picture is a description rather than an explanation.

• How old can a fungal individual get, and how much of a large genet stays connected?

  Why it matters: Both the age and size superlatives rest on extrapolation, and connectedness decides whether a genet is one organism or many clones.

• Can sequence-defined fungal taxa be reconciled with described species?

  Why it matters: Every estimate of fungal diversity depends on the answer, and the two methods currently count different things.