All techniques › Soil biology & amendments
Mycorrhizal inoculation
Adding bought mycorrhizal fungi to a soil that already has them. One field in four responds, a soil pathogen test rather than a fertility test predicts which, the drought claim rests on greenhouse work, and four in five products tested contained no live fungus. Cheap per hectare, and the free version is on other pages of this site.
The short answer
One field in four responds, and a fertility test will not tell you which. In 54 Swiss maize fields over three years, inoculation raised growth 12–40% in 14 fields and cut it 12% in two. Success followed the load of soil-borne fungal pathogens, not soil nutrients. On healthy, productive, well-fertilised fields the effect was zero or negative.
The drought claim is greenhouse evidence. Colonised plants in pots keep better water status. The two European experiments designed to test it, in Germany and Poland, found no yield benefit under drought. No European field trial exists.
Four in five products tested contained no live fungus. Of 25 commercial inoculants, 84% produced no colonisation in sterile soil and one raised plant growth. Demand a viability test on the batch, and start with the free version: less tillage, a winter cover and fewer fungicides each do more for the fungi you already have.
The one-in-four result first
The best European evidence is a trial run by the University of Zurich, Agroscope, the research institute FiBL and the University of Basel in 54 farmers' maize fields in northern Switzerland over 2018 to 2020 (864 plots). A well-characterised Swiss strain, Rhizoglomus irregulare SAF22, was worked by hand into the soil a few days after sowing (Lutz et al. 2023, Nature Microbiology, open access).
| Outcome | Fields |
|---|---|
| Significant gain, +12 to +40% dry biomass | 14 of 54 |
| Significant loss, −12% | 2 of 54 |
| No change or small decline | about a third |
| Range across all fields | −12% to +40% |
What separated the winners was not soil fertility but soil disease. The abundance of pathogenic fungi alone predicted a third of the variation; soil microbiome indicators plus a few chemistry values predicted 86%. In responsive fields the inoculant suppressed Olpidium, Cladosporium and Trichosporon. The follow-up found success correlated negatively with soil health and with the yield of uninoculated plots: "negative mycorrhizal inoculation success can occur in very healthy soils with high plant productivity" (FEMS Microbiology Letters, 2025). The authors call the one-in-four rate "the potential concern for the farmers" and see the product as a field-by-field decision guided by a soil diagnostic that does not yet exist as a service. Their hand application was "impractical for large-scale use due to high labor and costs"; clay granules through a microgranule spreader were on trial in 2024 (lab page).
What it is
Arbuscular mycorrhizal fungi colonise the roots of most crops and trade phosphorus, zinc and water for plant carbon. They are already in every European farm soil. "Inoculation" means adding more from a bag: spores or colonised root fragments of a few culturable species, almost always Rhizophagus irregularis (also sold as Rhizoglomus irregulare or Glomus intraradices), sometimes mixed with Funneliformis and Claroideoglomus species, bacteria or Trichoderma. Three things travel under the name and get different grades here:
- A bought product on field crops: in-furrow granule, seed treatment or soil drench.
- A bought product at planting for vines, trees and transplants raised in sterile substrate: the nursery use the products were designed for, and unassessed on this page.
- Managing for the native fungi: less tillage, cover crops, less phosphorus, fewer fungicides. That is the conservation-agriculture block already on this site.
How it is meant to help
Hyphae extend the effective root system into soil the root cannot reach, scavenging phosphorus and water and binding aggregates. Two things limit this in the field. The plant regulates the trade: on fertilised soils with ample phosphorus, crops cut the carbon they send to the fungus and the benefit falls to zero or below. And the native community is already there: an added strain must out-compete it, and establishes best exactly where the natives are weak, on high-phosphorus, disturbed, pathogen-loaded soils (Bender et al. 2019).
Whether farmers should manage for these fungi at all is disputed in the literature. Ryan and Graham (2018) found the field literature "overly optimistic" through "flawed methodology", with rigorous studies "only sometimes" showing a yield effect of high colonisation, "even under phosphorus limitation" (New Phytologist). Rillig and colleagues (2019) "could not support" that conclusion and argued for a lower-input systems view (New Phytologist). Neither side claims a bought product is a reliable input.
What the evidence shows
The meta-analyses vendors quote
| Synthesis | Scope | Result | Read with |
|---|---|---|---|
| Pellegrino et al. 2015 | Wheat, 38 field trials, 333 observations, 1975–2013 | Grain yield +20% with field inoculation (as reported) | "Economic profitability should still be addressed"; response driven by organic matter, pH, phosphorus, texture (abstract) |
| Zhang et al. 2019 | Seven cereals, 70 field-inoculation articles, English and Chinese | +16%, 77% of trials positive | Heavily Chinese and tropical; smaller in field than lab (abstract) |
| PeerJ 2022 | 21 rainfed field studies | +23% (16–30%) | Few or no European sites; drought not separated (open access) |
These pool laboratory inoculum with commercial products, tropical low-input soils with European ones, and researcher doses with farm ones. The European field trials are what a farmer here should read.
European field trials
| Where, when | Crop, product | Result | Source |
|---|---|---|---|
| Switzerland, 54 fields, 2018–20 | Maize, Swiss research strain | Gain in 14 fields, loss in 2, predicted by pathogen load | above |
| England, Leeds University farm, 2015–18 | Six wheat lines, commercial five-species mix, 35 kg N/ha | Ploughed 4.3 t/ha without inoculum, 3.5 with; disc 3.2 vs 3.3; inoculum "tended to reduce rather than increase mycorrhization"; "no tangible benefits to farmers or the environment" | open access |
| England (Elliott et al. 2021) | Three wheat cultivars, commercial inoculum, isotope tracing | Colonisation rose; uptake of mycorrhiza-acquired nutrients did not | abstract |
| Central Europe, Hungarian group, 4 years | Maize–wheat–maize–maize, three genera, Luvisol and Gleysol | Colonisation always higher; yield gain only with reduced fertiliser on the poor Luvisol, "not enough to compensate for the yield decrease caused by reduced fertilization" | abstract |
| Greece, Larissa, 2015–16, rainfed 266–316 mm | Barley, Czech product Symbivit | +9% seed yield on average (4.35 vs 4.25 and 4.33 vs 4.12 t/ha in 2015) | open access |
| Germany, Lower Saxony and Hesse, 2021 | Organic grain maize +15%, silage maize +31% | Vendor trials; no replication or statistics published | INOQ |
| Italy, Benevento, 2025 | Wheat, 3 ha treated vs one untreated plot, "+46.9%" | Vendor-adjacent; "not a randomized or replicated trial", not peer-reviewed | interim report |
The independent European results range from a loss to a small gain, with the Swiss quarter of fields as the best case. The large numbers all come from people selling the product.
Drought: greenhouse yes, field no
Two European experiments set out to test the drought claim. At Bielefeld (Germany, 2022) spring wheat in pots under continuous and pulsed drought with two commercial products showed a "negative mycorrhizal growth dependency" and "fungal presence could not alleviate the effects of drought" (open access). In Poland (2019) fodder maize recovered from a two-week drought faster with the fungus, but the effect on yield and quality "was negligible" (open access). Outside Europe, a semi-arid Jordanian field trial (2004) and a Moroccan one (2023, a native 15-species consortium plus compost, not a product) reported gains. No European field trial was found in which a bought inoculant raised yield in a drought year. The full case is on the claims page.
Is the fungus in the bag?
This is where the evidence is strongest, and it is against the products.
| Study | What was tested | Finding |
|---|---|---|
| Salomon et al. 2022, Applied Soil Ecology | 28 products, three continents; 25 in European and Australian greenhouse bioassays | Non-sterile soil: none raised colonisation, 1 of 25 raised biomass. Sterile soil: 84% produced no colonisation; laboratory R. irregularis colonised 48–79% (abstract) |
| Koziol, Lubin and Bever 2024 | 23 products | Propagule counts off by up to 100%; Olpidium pathogen contamination; no growth benefit once the fertiliser in the product is controlled for (abstract) |
| Koziol et al. 2025, New Phytologist meta-analysis | Commercial vs laboratory inoculants | 84% of commercial products under 5% colonisation; under 12% both colonised and helped, against 63% of lab cultures (as reported; paper) |
| Vahter et al. 2023, Mycorrhiza (Estonia) | 11 products from four producers, DNA sequencing | 9 contained at least one labelled species, 2 contained none, all but one contained undeclared species (abstract) |
| Boussageon et al. 2025 (Zurich) | 16 commercial products against 7 research cultures | Title: "Poor quality of commercial arbuscular mycorrhizal inoculants" (paper) |
Spores die in storage and heat, propagule counts are not viability tests, and fungicide seed dressings kill the inoculum on the seed. The claims page puts the numbers beside the label.
The free version
The fungi you already have respond to management, and the practices are on this site for other reasons. Across 54 field studies on five continents, reduced tillage and winter cover crops each raised colonisation of the following cash crop by nearly 30% against intensive tillage and winter fallow (Bowles et al. 2017). Across 150 cereal fields and 60 grasslands on a 3,000 km European transect, mycorrhizal phosphorus uptake was 64% higher in grassland than in cropland soils and fungicide use cut it by 43% (Edlinger et al. 2022, as reported). At Leeds the three-year ley reached 52% root colonisation against 36% under the plough, and 7.7 t/ha of wheat against 4.3. Colonisation is not yield, as Ryan and Graham insist, but the free version reliably delivers the first and costs nothing extra. See no-till, cover crops and ley–arable rotation.
Where it fits
| Context | Fit | Why |
|---|---|---|
| Field with poor soil health, known soil-borne fungal disease, low phosphorus, untreated seed, maize or legume | The case for trying it | The Swiss responders; cheap trial |
| Fertile, well-fertilised arable; healthy productive fields | No | Zero or negative in Switzerland, England, Hungary |
| Fungicide-dressed seed, brassica or beet rotations | No | Inoculum killed; non-host crops |
| Vines, orchards, transplants at planting | Unassessed | Designed use; no independent European field trial found |
| Rainfed Mediterranean cereals | One small positive | Greece +9%, one site, two years |
What it costs
| Item | Figure | Country, year | Status |
|---|---|---|---|
| In-furrow granule, 250 g treats about 17 ha | about US$30/ha | USA, 2025 | vendor and distributor, as reported |
| Licensed inoculant in the Benevento comparison | product €6.67/ha, applied €8.67/ha | Italy, 2025 | vendor-adjacent, unreplicated |
| Symbivit, 750 g garden pack | €23.97 (about €32/kg) | Germany, online retail, 2025 | not a field rate |
| RootGrow Professional, 5 L nursery pack | £80.42 ex VAT, 1,000–1,500 bare-root whips | UK, 2025 | planting use |
| Independent European cost per hectare, any route | none found | Pellegrino 2015: "economic profitability should still be addressed" |
Product cost for field crops therefore sits around €7 to €30 a hectare on vendor figures, plus application. Seed coating is cheapest but incompatible with fungicide dressings; in-furrow granules need a spreader on the drill; drenches need a sprayer pass. At those prices even the Greek 0.1–0.2 t/ha of barley pays in the years it appears. The problem is the three fields in four where nothing happens and the four products in five that may contain no live fungus.
Under the Fertilising Products Regulation (EU) 2019/1009 a CE-marked microbial plant biostimulant may contain only four groups, mycorrhizal fungi among them, and its claims must be substantiated under CEN technical specifications (EBIC summary). Nothing in that tests viability at the point of sale. Every figure here is also on the costs page. No entry in the stress test: there is no costed European margin gain to model.
Commitment
Fully reversible; the commitment is the annual invoice. No land change, no permit, no maintenance, no lasting effect for good or ill shown in European fields. The lowest-commitment technique in this library after mulching, which is not the same as a recommendation.
How to start
- Start with the free version. Cut tillage, keep a winter cover, avoid fungicide seed dressings where you can, and stop adding phosphorus the soil test says you do not need.
- Buy a product only for a field with a problem: poor soil health, known soil-borne fungal disease, low available phosphorus, a maize or legume crop, untreated seed.
- Demand proof of viability: a recent independent colonisation test on the batch, not a propagule count. If the supplier cannot provide one, the 84% figure applies.
- Run it as a farm experiment: paired strips with and without, same seed lot, same drill pass, weighed or yield-mapped, for two seasons. The trial is cheap; the annual habit is not.
- Store cold, use fast, keep away from fungicide, and do not apply to brassicas or beet.
Risks and failure modes
The bag is empty. Four in five products tested produced no colonisation. Invisible at purchase.
Nothing happens on good soil, and occasionally a loss: two Swiss fields, the ploughed Leeds plots.
Contamination with Olpidium and other pathogens detected in products.
Displacing native fungi without benefit: the strain can dominate roots while total colonisation and nutrient uptake stay flat.
Fungicide kills it, on the seed and in the crop.
Wrong host: brassicas, beet, spinach and buckwheat form no arbuscular mycorrhizas.
The reduced-fertiliser gamble. Cutting fertiliser in the expectation that the fungus compensates cost yield in the Hungarian trial.
Vendor arithmetic. Single-plot comparisons with 30–47% gains and net-profit projections are marketing. The peer-reviewed European range is −12% to +40%, with most fields at zero.
Who is talking. Producers, distributors, an institute that develops the products, the industry council and market-research firms dominate what a farmer finds. The independent European work, from Zurich, Agroscope, FiBL, Leeds, Tartu and the Hungarian group, is consistent and far more modest.
Claims checked. Mycorrhizal inoculants drought-proof your crops and The fungi on the label are in the bag.
Measured on real holdings. Two entries on the case studies page bear directly on this one: the 54-field Swiss trial network, where a significant gain appeared in 14 fields and a loss in two, and Spen Farm in England, where a bought inoculant did nothing in the same field where a three-year ley lifted wheat from 4.3 to 7.7 t/ha.