How to read this site
What the evidence grades mean, why commitment comes before cost, and why the same practice can be good advice in Ireland and bad advice in Andalusia.
Most writing about regenerative farming is confident. Techniques get recommended unconditionally, success stories circulate without measurements attached, and the honest caveats — this costs money, this only works in some climates, nobody has actually tested this at farm scale — tend to go missing.
This site is built the other way round. Every technique here was researched against peer-reviewed trials, meta-analyses and European field data, and the awkward findings were kept in. Several pages conclude "be careful". One or two conclude "the evidence does not support the advice you have probably been given."
That is the point. The weak grades are what make the strong ones worth believing.
The three badges
Every technique carries three labels, on its card and at the top of its page.
Evidence strength
How good is the research behind this, honestly?
| Grade | What it means |
|---|---|
| Strong | Multiple trials or meta-analyses, ideally including European data. |
| Good | Solid trial data, with gaps or questions about how well it transfers. |
| Mixed | Depends heavily on context. Read the conditions before acting. |
| Weak | Thin, modelled, or largely practitioner testimony rather than measurement. |
A weak grade is not a reason to dismiss a technique. It is a reason to trial it carefully rather than commit the farm to it — and to be sceptical of anyone presenting it as settled.
Where the most confident material on a subject comes from a body with a commercial interest — an input manufacturer, an industry association, or the inventor of the method — we say so on the page. That is not an accusation of dishonesty. It is context you are entitled to.
Commitment
We show this before cost, deliberately. Price is easy to find out; what a practice locks you into is not, and it is usually the more consequential question.
| Level | What it means |
|---|---|
| Reversible | Stop after one season at no lasting cost. |
| Moderate | Some capital or a few years to unwind. |
| Long | Years to decades — but a mild downside if abandoned. |
| Permanent | Irreversible, with an ongoing maintenance obligation. |
Note that long commitment and high risk are not the same thing. Agroforestry is the longest commitment here, but an abandoned agroforestry plot becomes woodland and the timber holds value. An abandoned terrace collapses and causes worse erosion than never building it. Length and severity are separate axes and we treat them separately.
Cost
Free or near-free, low, medium or high. Where we have European figures we give them with the country and year attached, because prices and policy move — a straw price from 2019 tells you very little about 2026.
Where we only have non-European figures, we say that rather than converting them and hoping. Several pages therefore have thinner cost sections than we would like. That is an honest gap, not an oversight.
The free version is usually the better-evidenced one
This was not a position we set out with. It is what the research kept finding. In six of the seven pairings we examined in depth, the free or on-farm form of a technique turned out to have better evidence behind it than the version with an invoice attached.
| The free or on-farm form | The paid form | What the evidence says |
|---|---|---|
| Mulching with your own pruning residue | bought mulch or a soil conditioner | The strongest evidence grade in this library, at no material cost |
| Managing for native mycorrhizae through less tillage and a winter cover | a bought inoculant | Each raises root colonisation by nearly 30% across 54 field studies, while 84% of 25 tested products produced no colonisation at all |
| Free regional irrigation scheduling services | probes and a platform subscription | Emilia-Romagna's free service saves 90 million cubic metres a year, at about €0.02 a hectare of platform cost |
| Deficit irrigation as a strategy | a sensor network | Over 20% of water saved across five Spanish olive orchards with no loss of yield or oil quality, using no hardware at all |
| Free public drought monitoring | a commercial imagery platform | The platforms resell the same free satellite imagery, and no European trial was found of one changing a farm decision for the better |
| Compost made from farm residues | biochar or hydrogels | Compost has European farm-account evidence; biochar costs €4,000–18,000 a hectare at the rates trials use; hydrogels have three European field trials and a compliance deadline in 2028 |
One clear exception, and it matters. Buried drip irrigation is expensive engineered hardware, and it is graded good on this site: roughly 55% less water than furrow or flood irrigation, maize yields up 31%, nitrogen efficiency up 43 to 71%. The cheap traditional alternative on the same page, buried clay pots, is useful in a market garden and does not scale to a field, and the claim that they are ten times more efficient than sprinklers traces to people selling them. Here the paid option is plainly the better-evidenced one. We put that in the table's way deliberately, because a rule with no exception is usually a belief.
Why the pattern keeps appearing is not a conspiracy. A practice nobody can patent is a practice nobody funds trials of, so the free options accumulate evidence slowly, from public bodies and university trials. A product that can be sold funds its own trials, and those are the figures that reach a farmer first. The result is not that vendors are lying. It is that the best-evidenced version of an idea is often the one with no marketing budget behind it, and you have to go looking for it.
What to do with this. Treat the free version as the first thing to try, not the fallback. On six of the seven pairings we have examined this way, it was the right first move, and the seventh is buried drip. That is a claim about where the evidence currently sits, and it would change if someone ran the trials.
The 700 mm line
The single most important number on this site is roughly 700 mm of annual precipitation.
Below that threshold, several standard regenerative recommendations reverse. The clearest case is cover crops: a growing cover crop transpires water, which is harmless in a wet climate where rain replaces it, and directly harmful in a dry spring where it comes out of the reserve your cash crop needed. Trial data shows the following crop gaining 4% in continental Europe and losing 11–12% in dry and temperate-dryland conditions.
Much of Mediterranean and south-eastern Europe sits below that line — which is precisely where the drought problem is worst. So the practices most enthusiastically promoted as drought solutions are sometimes the ones that need the most caution in the driest places.
This is why the technique library filters by climate zone, and why some techniques do not appear under Mediterranean at all. The filter is telling you something, not hiding options.
If you would rather answer a few questions than work the filters, the decision tool re-ranks every technique for your rainfall, soil, tenure, system and equipment, and shows the reason for every demotion. Nothing is hidden; the point is the reasons.
Does it work, and does it pay
These are two separate questions and this site answers both, because a practice can improve your soil and still lose you money.
The clearest example again comes from cover crops. UK trials found gross margin reduced in 95% of situations, averaging −£150/ha across two consecutive cash crops — in a climate where the agronomy is favourable. The reason was simply that the yield benefit did not cover establishment costs.
Subsidy does not automatically close such gaps either. The only European eco-scheme increment we have been able to verify is €20/ha, which the researchers who examined it judged insufficient to offset the cost of changing practice.
None of that means don't do it. It means do it for a reason you can name: a scheme payment that genuinely pays, an erosion or leaching problem already costing you money, livestock that can use the cover. "It improves the soil" is true and is not, on its own, a business case we can support with the evidence.
Reduce demand before improving supply
A pattern emerged across the water techniques that is worth stating as a principle. The practices worth trying, in order:
- Store water in the soil. Organic matter and surface cover have no evaporation penalty, need no permits, and cost little. Mulching is the obvious start.
- Reduce how much water the crop needs. Shelter from trees cuts evapotranspiration by 10–30% over a zone many times the height of the trees. Reducing demand beats competing for supply.
- Deliver irrigation efficiently, below the surface where it does not evaporate.
- Store water in the open, last. Measured evaporation from a Portuguese reservoir ran to 450 mm across a single summer, around 45% of the depth of a one-metre pond.
There is a catch at step three worth knowing before you invest. Making irrigation more efficient has repeatedly failed to save water at regional scale, because the saving gets spent on expanding irrigated area or switching to thirstier crops. Spanish drip irrigation expanded 76% between 2001 and 2019 while total consumptive water use rose. Efficient irrigation will help your farm; it will not by itself help your region, and if you bank the saving as extra hectares you have made your own basin's drought worse.
What this site is not
It is not a certification scheme, a funding application, or a substitute for local advice. Soil type, slope, rotation, machinery and tenure all change the answer, and a page cannot know those. Where a practice needs a permit — earthworks and water interception usually do — we say so, but the rules are national and you need to check yours. The permits page sets out England, Spain, Poland and Portugal side by side.
It is also incomplete, and visibly so. The technique library now covers every practice we found real European evidence for, so it grows in depth rather than breadth; what is still thin is cost and funding data outside a handful of countries, and the costs page names those gaps one by one. We would rather publish a smaller library properly than a large one superficially.
Where the numbers come from
Sources are linked inline on every page, so you can check any figure yourself rather than taking our summary of it. We prefer peer-reviewed studies, meta-analyses, European field trials, and material from the EEA, FAO and Copernicus. Where the best available evidence is non-European — a lot of the mulching literature comes from Asian drylands — we say that too, along with how well we think it transfers. Where a European farm or farm-scale site actually measured a technique over years, it is on the case studies page, together with the candidates we rejected and why.
Where a confident claim kept turning up while we researched, we ran it down to its source; the results are on the claims page, each with a verdict on the sentence as it circulates rather than on the practice.
If you find something here that is wrong, it is wrong in public and worth correcting. Two pages already carry corrections to earlier conclusions we reached, because better evidence turned up later.