On mycorrhizal fungi, the underground networks that feed your most beloved plants, and why some flowers simply cannot bloom without them
Beneath every thriving garden, there is another garden entirely β one you will never see with the naked eye, one that operates in darkness and silence, and one that your most spectacular blooms depend upon absolutely. It is a garden of fungi: thread-thin filaments threading through soil particles, latching onto root tips, exchanging minerals for sugars in transactions so ancient that the first land plants on Earth could not have colonised the terrestrial world without them.
The relationship between plants and fungi is one of the oldest partnerships in the history of life. It predates flowers. It predates trees. It predates, by several hundred million years, the emergence of any animal capable of admiring a garden. And yet most gardeners tend their beds in complete ignorance of it β feeding, watering, pruning, deadheading, all the while unknowingly disrupting or sustaining a microscopic economy that determines, more than almost any other factor, whether certain plants will merely survive or genuinely flourish.
This guide is about those plants β the ones that are not simply aided by fungal relationships but are, in a meaningful sense, built around them. Orchids that cannot germinate without specific fungal partners. Heathers that struggle on ordinary soil because they evolved on fungal networks in acidic moorland. Peonies and roses that reward you with their largest, most fragrant blooms when the soil beneath them is alive with mycorrhizal threads. Trilliums and other woodland wildflowers that are so deeply embedded in the underground web of the forest floor that transplanting them without care for what lives in the soil is simply a slow way of killing them.
Understanding these relationships does not require a background in mycology. It requires only a willingness to think of soil as something other than a medium for anchoring roots β to think of it instead as a community, with its own economy, its own communication networks, its own long history.
What Mycorrhizal Fungi Actually Are
The word mycorrhiza comes from the Greek for fungus (mykes) and root (rhiza). It describes the symbiotic union between a fungus and the root of a plant β a union so intimate that, under a microscope, it is sometimes difficult to say where one organism ends and the other begins.
The relationship works, in its simplest form, like this: the fungus extends far beyond the reach of the plant’s roots, mining the soil for phosphorus, nitrogen, zinc, copper, and other minerals that the plant cannot access efficiently on its own. In exchange, the plant feeds the fungus with carbohydrates produced through photosynthesis β sugars that the fungus, being unable to photosynthesise, cannot make for itself. It is a trade, conducted in chemical currency, that benefits both parties.
There are two main types of mycorrhizal fungi, and understanding the difference between them matters for gardeners.
Ectomycorrhizal fungi form a sheath around the outside of root tips without penetrating the root cells themselves. They are associated primarily with trees β oaks, beeches, pines, birches β and they include the familiar woodland mushrooms: penny buns, chanterelles, truffles, fly agarics. When you find a cep growing beneath an oak, what you are looking at is the fruiting body of a fungus whose threads are intertwined with the roots of that tree. The mushroom is the visible tip of an invisible partnership.
Arbuscular mycorrhizal fungi (AMF), by contrast, penetrate directly into root cells, forming branching structures called arbuscules inside the cell walls β the actual site of nutrient exchange. They are associated with the vast majority of flowering plants, including most vegetables, grasses, and garden perennials. They produce no large fruiting bodies; you will never find an AMF mushroom in your border. Their entire existence is underground and microscopic.
Both types are destroyed by synthetic phosphorus fertilisers, because the fungus forms its partnership with a plant that is searching for phosphorus. A plant that has been fed high-phosphorus fertiliser has no need of the fungal partnership and ceases to maintain it. This is one reason why heavily fertilised, pesticide-treated soil tends to be mycorrhizally impoverished β and why plants that evolved in fungal-rich conditions struggle in such soil even when supplied with adequate nutrition by conventional means.
Orchids: An Absolute Dependency
No group of plants demonstrates the necessity of fungal partnership more dramatically than orchids. In the orchid family, the relationship with fungi is not merely beneficial β it is, for most species, an existential requirement.
An orchid seed is the smallest seed produced by any flowering plant. A single seed pod can contain hundreds of thousands of them, each one a tiny parcel of genetic material wrapped in a gossamer coat, with almost no nutritional reserves whatsoever. Compared to, say, a bean seed β which contains enough stored starch to fuel germination and push a seedling toward the light before it becomes capable of photosynthesis β an orchid seed is almost entirely empty. It cannot germinate without an external energy source.
That energy source is a fungus.
When an orchid seed lands on suitable soil or bark, it must make contact with the right fungal species within days or it will die. The fungus penetrates the seed and begins to feed it, essentially digesting itself to fuel the embryo’s development. This is a parasitic phase β the orchid is taking from the fungus without giving anything back β and it can last for years in species that have a slow germination and development process. Only later, once the orchid has developed chlorophyll and can photosynthesise, does the relationship shift toward mutualism.
In terrestrial orchids β the wild species that grow in meadows, woodland clearings, and heathland β the adult plant often continues to depend on the fungal partner throughout its life, drawing on fungal carbon during periods when its own photosynthesis is insufficient. Some terrestrial orchids, most famously the ghost orchid (Epipogium aphyllum) and the bird’s nest orchid (Neottia nidus-avis), have taken this dependency so far that they have lost their chlorophyll entirely. They are fully mycoheterotrophic β obtaining all their nutrition from the fungal network, contributing nothing in return. They are, in effect, plants that have evolved into sophisticated fungal parasites.
For gardeners, this has profound practical implications. Wild terrestrial orchids β green-winged orchids, bee orchids, pyramidal orchids, early purple orchids, spotted orchids β cannot simply be planted into a bed. They need the specific fungal partners present in the soil of their native habitats, and they need the plant communities that sustain those fungi. The best way to establish native orchids in a garden is to introduce plugs of soil and turf from an established orchid meadow, or to purchase plugs that have already been inoculated with the correct fungal species. Even then, success is not guaranteed, because the fungi themselves have specific requirements about soil chemistry, moisture, and companion plant communities.
Tropical epiphytic orchids β the Phalaenopsis, Cattleya, and Dendrobium grown on windowsills and in glasshouses β have different relationships with their fungal partners, but still benefit significantly from bark-based growing media that allow fungal colonisation. The use of conventional potting compost for orchids is almost universally detrimental, not only because of drainage issues but because the compost environment disrupts the fungal communities the plant evolved alongside.
Heathers and Heathland Plants: The Ericoid Network
Walk across a Scottish moor in August and the ground beneath the heather, though it looks like nothing more than saturated peat, is threaded with a specific type of mycorrhizal fungus found almost nowhere else. Ericoid mycorrhizal fungi β so called because they are associated with the Ericaceae family, which includes heathers, blueberries, rhododendrons, azaleas, and pieris β have evolved to operate in conditions that would be hostile to most fungi: highly acidic, low in nutrients, and rich in organic compounds that are difficult to break down.
What ericoid fungi do, with unusual efficiency, is break down these complex organic molecules and extract nitrogen from them directly β a feat that most other fungi and the plants themselves cannot manage. In the nitrogen-poor environments of heathland, bog, and acid woodland, this ability is transformative. Heather (Calluna vulgaris and Erica species) grown in poor, acid soil with active ericoid mycorrhizae will thrive; the same heather in a standard garden bed, with the soil’s fungal community disrupted by cultivation and conventional fertilisers, will sulk, yellow, and eventually fail.
This explains one of the most common puzzles in domestic horticulture: why ericaceous plants so often struggle in even dedicated ericaceous compost, producing poor growth and sparse flowers despite what seems like appropriate soil conditions. Commercial ericaceous compost provides the correct pH but is typically sterile β heat-treated to destroy pathogens and weed seeds, but in the process stripped of the mycorrhizal fungi the plants evolved with. The soil is the right acidity; it simply contains none of the biology.
The practical solution is to inoculate ericaceous plantings with ericoid mycorrhizal inoculants at the time of planting, and to mulch with composted bark or pine needles rather than ordinary compost. The mulch serves as a substrate for the fungal community; the inoculant establishes the partnership. Plants treated this way β rhododendrons, blueberries, azaleas, pieris, Enkianthus β consistently produce more vigorous growth, better flower set, and more intense colour than unfungalised equivalents grown in the same soil.
Blueberries are a particular case worth dwelling on. The ericoid mycorrhizal network is essential not just for their growth but for the flavour and nutritional density of the fruit. Studies comparing blueberries grown with active ericoid fungal associations to those grown in mycorrhizally impoverished soil have found meaningful differences in anthocyanin content β the compounds responsible for both the deep blue colour and a significant proportion of the fruit’s health benefits. You are not just growing more blueberries; you are growing better ones.
Roses: Ancient Partners, Modern Neglect
The rose has been cultivated for so long, and subjected to so much intensive breeding, chemical treatment, and horticultural interference, that it might seem like a plant that has moved beyond its ecological origins entirely. And yet beneath a thriving rose bush, if the soil has been left undisturbed and free of synthetic fertilisers, arbuscular mycorrhizal fungi are present and active β and their presence makes a measurable difference.
Studies on mycorrhizal colonisation in Rosa species consistently show that roses with active AMF associations develop larger, more ramified root systems, access phosphorus and zinc more efficiently, show significantly greater resistance to drought stress, and produce blooms more abundantly and with higher essential oil content. That last point is not trivial for anyone who has ever wondered why old roses from garden centres often lack the fragrance that antique varieties are celebrated for β fragrance in roses is partly genetic, but it is also significantly influenced by the availability of trace minerals, particularly zinc and boron, both of which are delivered more efficiently via mycorrhizal pathways than through direct root uptake.
The modern practice of rose cultivation has been, in many respects, systematically hostile to these partnerships. Conventional rose fertilisers are typically high in phosphorus; fungicidal sprays β applied routinely to combat black spot, rust, and mildew β do not discriminate between pathogenic and beneficial fungi; and the habit of digging and cultivating rose beds annually disrupts the hyphal networks that take months to establish. The result, in many gardens, is roses that depend entirely on chemical inputs to perform at all, because the biological systems that would otherwise support them have been eliminated.
The shift toward mycorrhizal-aware rose cultivation involves relatively simple changes: using fertilisers low in phosphorus, applying mycorrhizal inoculant at planting time, mulching with well-composted wood chip rather than tilling the soil, and accepting that biological fungal control will always be preferable to chemical fungicide. Roses treated this way take longer to establish than those pushed with conventional feeding, but they become, over two or three seasons, more self-sufficient, more fragrant, and notably more resilient to the diseases that beset their over-fertilised counterparts.
Peonies: Deep Roots, Older Networks
The peony is, as any gardener who has tried to establish one will know, a plant with strong opinions about where and how it lives. It dislikes being moved. It resents disturbance. It can sulk for two or three years after transplanting before it condescends to flower again. These behaviours, which can seem like mere horticultural temperament, make considerably more sense when understood in the context of mycorrhizal dependency.
Peonies form extensive AMF associations, particularly with fungi in the Rhizophagus irregularis complex β one of the most widespread and studied of the arbuscular mycorrhizal species. These associations take time to establish and are highly sensitive to soil disruption. When a peony is moved, it loses not just its established root system but its entire fungal network β the threads that have been, over months and years, extending through the surrounding soil and connecting the plant to a complex underground economy. Rebuilding that network from scratch is the real reason a transplanted peony takes so long to recover.
This also explains why peonies respond so dramatically to the quality of their soil biology. A peony planted into soil with an active, diverse mycorrhizal community β old garden soil, woodland edge soil, or a bed that has been mulched with composted wood chip for several years β will typically outperform a peony planted into fresh topsoil or commercial compost, even if the conventional agronomic analysis of the two soils would favour the latter. The chemical richness of the soil matters less than its biological richness.
For gardeners wishing to give a new peony the best possible start, the most important intervention is not fertiliser but inoculant: a granular or gel mycorrhizal product applied directly to the root ball at planting, so that the fungal partnership begins immediately rather than waiting for natural colonisation from the surrounding soil. In a newly cultivated bed β where soil disturbance and possibly previous chemical use have impoverished the fungal community β natural colonisation may be slow or partial. Inoculation shortcuts that process significantly.
Woodland Wildflowers: The Forest Floor Economy
In a mature deciduous woodland, the mycorrhizal network beneath the leaf litter is not a collection of isolated partnerships between individual trees and fungi. It is a web β interconnected, overlapping, shared between dozens of species, mediating the flow of nutrients and carbon across an area of forest floor that may extend for hectares. This network has been described, in popular science writing, as the wood wide web β a term that, for all its anthropomorphising simplicity, captures something true about the interconnectedness and information-sharing capacity of the system.
Many of the most beautiful woodland wildflowers β trilliums, wood anemones, bluebells, wild garlic, Solomon’s seal, lily of the valley, hepaticas, bloodroot β depend on this network not only for nutrition but for carbon during the months when they are dormant or before their leaves have fully deployed in spring. They are not simply aided by the fungal web; they are participants in it, drawing on shared resources and contributing to them when they can.
This dependency is why woodland wildflowers are so notoriously difficult to establish in garden settings. The problem is not soil chemistry β most of these plants are relatively unfussy about pH β but soil biology. A garden border, however well-prepared, rarely contains the complex, species-rich mycorrhizal community of a woodland floor. The plants that depend on that community for their earliest weeks of spring growth, before their leaves are working at full capacity, simply do not thrive without it.
Trilliums are perhaps the most extreme example. Trillium grandiflorum and its relatives take seven years or more to flower from seed in the wild β a timeline that is partly a function of slow germination but largely a reflection of the time required to build a functional mycorrhizal partnership and accumulate sufficient carbon reserves. Trilliums purchased from nurseries that have sourced them through wild collection β a common and damaging practice β are often torn from fungal networks that cannot be recreated in a garden, and they frequently fail within a season or two despite appearing healthy at the point of purchase. Nursery-raised trilliums, grown with their fungal partners from the beginning, are very different plants.
Bluebells (Hyacinthoides non-scripta) form AMF associations that are integral to their extraordinary performance in British ancient woodland β the ability of millions of bulbs to mobilise nutrients simultaneously in the brief window between leaf-break and full canopy closure. A bluebell planting in a garden, in soil without an established woodland fungal community, will typically grow well enough but rarely achieves the density and intensity of colouration seen in genuinely ancient bluebell woods. The difference is biological, not horticultural.
Hepaticas (Hepatica nobilis and related species) are among the most coveted of woodland wildflowers β the early-spring flowers in shades of blue, violet, pink, and white β and among the most reliably difficult to establish. Their relationship with ectomycorrhizal fungi of the forest floor is subtle but important; in the right soil biology, with the right tree cover, they can be extraordinarily long-lived and generous in self-seeding. In the wrong soil, they persist reluctantly for a season or two and then vanish. The difference in outcome is rarely explained by any factor visible above ground.
Trees: The Ectomycorrhizal Canopy
Above the herbaceous layers of the garden, the trees carry their own fungal partnerships β and those partnerships are, at larger scales, the source of the mycorrhizal network that woodland wildflowers depend upon.
Oak trees (Quercus species) form associations with over two hundred species of ectomycorrhizal fungi β a diversity of partnership that reflects millions of years of co-evolution and explains much of the ecological richness associated with ancient oak woodland. A mature oak is not just a tree; it is a hub in a fungal network that extends through the soil to neighbouring trees, shrubs, and herbaceous plants, mediating nutrient flows across the entire community.
Pine trees similarly form deep ectomycorrhizal associations, particularly with fungi in the Suillus and Rhizopogon genera. Young pine seedlings that germinate without appropriate mycorrhizal partners are noticeably smaller, paler, and less vigorous than inoculated equivalents β a difference that becomes more pronounced, not less, as the trees mature and their dependence on the fungal network for nitrogen deepens.
Birches (Betula species) are notable for the speed and breadth of their mycorrhizal colonisation β they tend to be among the first trees to establish in disturbed ground, partly because they are mycorrhizal generalists capable of partnering with a wide range of fungal species, and partly because they are active participants in the fungal network-building that subsequently allows more specialist woodland species to establish.
For gardeners planting trees, the practical implication is the same as for herbaceous plants but with a longer time horizon: mycorrhizal inoculant applied at planting is one of the highest-return investments you can make in a young tree’s future, particularly in disturbed or cultivated ground where the native fungal community has been impoverished.
What Destroys Mycorrhizal Networks β and What Restores Them
Understanding which plants benefit from mycorrhizal partnerships is only half the picture. Equally important is understanding what disrupts those partnerships in garden soils, and what can be done to restore them.
The main disruptors are familiar to most gardeners, though their effects on soil biology are less often discussed than their effects on pests and diseases.
Synthetic phosphorus fertilisers are the single most damaging input for mycorrhizal communities. Phosphorus is the currency of the mycorrhizal exchange; a plant that has abundant phosphorus supplied directly to its roots has no metabolic incentive to maintain its fungal partnerships and actively suppresses them. Soils that have received high-phosphorus fertilisers for several seasons can take years to recover their mycorrhizal community even after fertilisation stops.
Fungicides β including both systemic fungicides applied as soil drenches and contact fungicides applied to foliage β inevitably affect non-target fungi in the soil. Broad-spectrum fungicides are particularly damaging; even products marketed for specific pathogens have demonstrated effects on mycorrhizal species in controlled studies. The cumulative effect of routine fungicide use in a garden border can be a significant impoverishment of the soil’s fungal diversity over time.
Rotavating and deep digging physically sever hyphal networks that may have taken months or years to establish. Mycorrhizal threads are extraordinarily fine β many are thinner than a human hair β and extremely fragile. The kind of thorough soil cultivation that was once considered best horticultural practice is, from the perspective of soil biology, more like a natural disaster.
Bare soil, left uncovered between plants, loses its fungal community gradually through UV exposure, desiccation, and the absence of the plant roots that sustain the fungi. Mulching β with wood chip, leaf mould, or composted bark β protects the soil surface and provides both the carbon substrate and the moisture that fungal communities require.
The main restorers are, by comparison, simple.
Reducing or eliminating synthetic phosphorus fertilisation, and substituting slow-release organic fertilisers that deliver phosphorus gradually, allows the mycorrhizal community to re-establish over one to three growing seasons.
Mycorrhizal inoculant products β available from most good garden centres and specialist suppliers β provide direct introduction of fungal spores and propagules to new plantings. They are not a substitute for a healthy soil biology, but in impoverished soils they can significantly accelerate the process of recovery. The best products are applied directly to the root ball or root zone at planting, not to the soil surface.
Companion planting with species that sustain active mycorrhizal communities benefits neighbouring plants through the shared network. Planting meadow grasses, wildflowers, and deep-rooted perennials in proximity to mycorrhizally dependent plants creates the community of roots that the fungal network requires to sustain itself.
The no-dig approach to garden cultivation β leaving soil structure undisturbed and working with mulch and compost applied to the surface β is now well-established as a method that produces healthier, more biologically active soil over time. From a mycorrhizal perspective, it is simply the most rational way to garden around organisms whose architecture is built on fragile, permanent threads.
A Summary of Key Plants and Their Fungal Needs
Orchids (all terrestrial species; most epiphytic species): Absolute dependency for seed germination; continued dependency in many species throughout life. Require specific fungal partners; inoculant products are available for some species, but habitat recreation is the most reliable approach.
Heathers and ericaceous plants (Calluna, Erica, rhododendrons, azaleas, blueberries, pieris): Dependent on ericoid mycorrhizal fungi for nitrogen uptake in acid conditions. Require ericoid-specific inoculants; standard AMF products are not appropriate.
Roses: Strongly benefit from AMF associations for phosphorus and zinc uptake, drought resilience, and fragrance compound production. Use AMF inoculant at planting; avoid high-phosphorus fertilisers and broad-spectrum fungicides.
Peonies: Deep AMF dependency; highly sensitive to soil disturbance and chemical disruption. Use AMF inoculant at planting, minimise soil disturbance, and mulch generously.
Woodland wildflowers (trilliums, bluebells, wood anemones, hepaticas, Solomon’s seal, lily of the valley): Depend on established woodland mycorrhizal networks. Most reliably established by introducing plugs of woodland turf, or purchasing nursery-raised specimens with intact fungal associations.
Trees (oaks, pines, beeches, birches): Form ectomycorrhizal associations essential for long-term vigour and resilience. Use ectomycorrhizal inoculant at planting; avoid cultivation around root zones of established trees.
Lavender and Mediterranean herbs: Benefit significantly from AMF associations, which explain their preference for poor, well-drained soils β conditions that favour mycorrhizal activity over direct root uptake.
Alliums (ornamental and culinary): Active AMF associations; one reason that over-feeding alliums with nitrogen fertiliser tends to produce lush foliage at the expense of flower and flavour.
Prairie and meadow perennials (Echinacea, Rudbeckia, Baptisia, native grasses): Highly mycorrhizal; the reason that meadow plantings established in undisturbed or minimally cultivated soil consistently outperform those planted into deeply prepared beds.
The Garden Beneath the Garden
There is a shift that happens in a gardener’s relationship to their soil when they begin to understand what lives in it. The border stops being a collection of plants arranged in more or less appealing combinations and becomes something more like a community β one in which the visible part, the part above ground, is in some ways the least interesting portion.
Below the surface, in the dark, in the cold, in the wet, an economy older than any flower is conducting transactions of extraordinary complexity. Nutrients moving from mineral deposits to fungal threads to root tips. Carbon flowing in the other direction. Information, in the form of chemical signals, travelling through hyphal networks from plant to plant. Stress signals. Resource-sharing. Something that, if pressed to name it without too much anthropomorphising, you might call cooperation.
To garden in a way that supports this economy β reducing chemicals, minimising soil disturbance, inoculating at planting, mulching the surface, welcoming the full complexity of what lives in the ground β is not an act of sentimentality. It is a recognition that the most beautiful flowers in your garden did not get there by themselves. They had help from below.
They always have.