The EU orders its soil fungi measured, not graded

A line in Annex I

Directive (EU) 2025/2360 on soil monitoring and resilience is dated 12 November 2025, appeared in the Official Journal on 26 November 2025 and entered into force on the twentieth day after that, 16 December 2025. Its first annex carries a table headed "Part C: soil descriptors without criteria", and against an aspect of soil degradation called "Loss of soil biodiversity" it sets a single obligatory descriptor: "DNA metabarcoding for fungi and bacteria". You can read the act itself on EUR-Lex in every official language of the Union.

One table row, but a real obligation behind it: member states will have to take soil samples and read fungal DNA sequences out of them. The fungi themselves occupy very little of the text. The stem "fung" occurs exactly four times in the whole directive — three of them in recital 21, the fourth in that annex row. Recital 21 notes that soils host more than 25 per cent of all biodiversity, mentions mycorrhizal fungi connecting plant roots, and concludes that the collection and analysis of information on soil bacteria and fungi should be recognised and should "serve as a foundation for the potential future expansion of biodiversity monitoring". That is a good deal weaker than a conservation objective. Fungi did not cause this law; they are something it has started to count.

How much the directive actually requires

Not every sampling point will be sequenced. Article 9(3), in its sixth subparagraph, requires measurements on "at least 5 % of the total number of sampling points" determined under Article 9(1). How many points that fraction represents is written nowhere in the act. Under Article 9(1) and Annex II, Part A, each member state designs its own stratified random sample, stratified on soil units, so that the points "represent the variability of the chosen soil descriptors within the soil units, with a maximum error percentage (or coefficient of variation) of 5 %". The Commission is to hand each country maps of the descriptors, initial sampling points and the data attached to them from earlier LUCAS soil surveys. The number of European soil samples that will actually be metabarcoded therefore cannot be stated until national sampling plans exist.

The timetable is long. National law must be in force by 17 December 2028, the first soil measurements must be performed by 17 December 2030 and repeated every six years, and the first soil health assessment is due by 17 December 2031. That does not mean nothing counts yet. Article 9(4) lets data gathered from 16 December 2024 onwards feed the first cycle, drawn from existing national or subnational monitoring networks, from measurements made under Union and international law, and from private actors and research organisations. The careful statement is therefore that no result is owed under this directive before 2030, not that no relevant fungal data is accumulating. It is also a directive rather than a regulation: it works through national statutes that have yet to be written, and the detail of the methods can differ between them.

Measured, but not marked

The most interesting limit follows from where the descriptor sits. Article 7(1) does tell member states to apply the descriptors in Parts A, B and C when monitoring and assessing soil health. But the criteria for healthy soil condition in Article 7(2) are the non-binding sustainable target values in Parts A and B, together with operational trigger values that Article 7(6) provides only for Parts A and B. Article 10(1) is blunter still: soil health is assessed "for each of the soil descriptors listed in Annex I, Parts A and B". The fungal figure will not feed the test that decides whether a given soil counts as healthy or as degraded.

It is not simply filed away, either. Article 10(3) obliges member states to analyse the Part C values "with a view to identifying whether there is a critical loss of ecosystem services, taking into account the relevant data and available scientific knowledge". That is an analysis without a ruler, because the act supplies no value to measure the result against, and Article 10(4) adds only that member states "may identify improvements" for descriptors in Parts A, B and C. There is no threshold below which a fungal community counts as impoverished, and no duty to rebuild one.

This is not settled for good. A footnote to Annex I contemplates target values for Part C as well, "when possible", and Article 25 requires the Commission to evaluate the directive by 17 June 2033 and to weigh both the establishment of criteria for Part C descriptors and a higher percentage of sampling points for the soil biodiversity descriptors. There is no threshold now; that is a different statement from saying there will never be one.

A second gap concerns method. In the methodology table of Annex II, Part B, every other descriptor carries a named standard — ISO 11277 for texture, ISO 10390 for acidity, ISO 11272 for bulk density. For the row "Soil descriptors linked to soil biodiversity and biological activity" the reference-methodology cell is empty. The minimum methodological criteria amount to one sentence: European or international standards shall be used where available, and where no such standard exists the chosen methodology must be available in the scientific literature or publicly available. The last column, which asks whether a validated transfer function is required, reads "Not applicable" — because there is no reference method to convert anything to. The word "metabarcoding" appears twice in the entire act, both times in Annex I, and not once in Annex II. Primer choice, sequencing depth and bioinformatic processing can therefore differ between member states, and the directive does not guarantee that twenty-seven national datasets will be comparable with one another.

It is worth remembering what metabarcoding does. It detects DNA sequences present in a sample. It does not count individuals, weigh mycelium or assess the condition of a fungus; it reports whose DNA was in a few dozen grammes of soil. That is useful and narrow, and the act is right not to ask more of it than it can give.

Where the figures come from, and what to watch

A number attached to this law is not in it. The claim that over 60% of European soils are in an unhealthy condition comes from the Council of the EU press release of 29 September 2025, which attributes it to available data. The directive's own recital 2 estimates something different: that "60 to 70 % of the soils in the Union are degraded and continue to deteriorate". Two figures, two wordings, and they should not be merged.

The same release invites a dating error. Its headline says the Council adopted the rules, but what it describes is the Council's position at first reading, and the page itself expects Parliament's final vote in the coming weeks. A footnote in the directive records Parliament's position of 23 October 2025; signature followed on 12 November, publication on 26 November, and the Commission's Mission Soil platform confirms entry into force on 16 December 2025 with a three-year transposition deadline.

Nor is this the first reading of fungal DNA from European soil. For the biodiversity module of the LUCAS Soil 2018 survey the Commission's Joint Research Centre sequenced bacterial 16S and fungal ITS from 885 soil samples and released the raw reads. What changes is who owes the work and how often: LUCAS is a Commission survey, whereas the directive puts the measurement on the member states and makes them repeat it every six years.

The next document worth watching is consequently national rather than European. Each member state must publish its transposing law by 17 December 2028, and those texts will answer what the directive leaves open: who takes the samples, how many points beyond the minimum are sequenced, and by which method the fungal DNA is read. Those choices, not the annex, will decide whether the first cycle produces a comparable picture of what lives beneath Europe's fields and woods, or merely a set of numbers that satisfies a floor. Anyone can begin checking today by opening Annex I and counting how many rows describe the chemistry of a soil and how many describe its life.