A versioned agronomy protocol, published openly, reported back on.
Not a certification and not a standard. A written method — drilling dates, rotations, margin widths, input thresholds — with a version number on it, that any farm can adopt, and that only improves if the farms that adopt it send back what happened.
The protocol, specified.
What it is
- A written method
- Versioned, dated, and publicv1.3 is a different document from v1.2 and both remain available for ever.
- Scope
- Combinable crops on thin calcareous soilsDeliberately narrow. A method that claims to work everywhere works nowhere.
- Licence
- Open, attribution, no field-of-use restrictionA farm can adopt it, adapt it, ignore it or fork it.
- Cost
- Nothing, and no membershipA protocol with a fee is a certification scheme, which is a different and much worse thing.
What it contains
- Drilling windows, by crop
- With the yield and blackgrass trade-off statedThe single most contested decision, written down with both halves of the argument.
- Rotations
- Three, for three levels of weed pressureNot one rotation. Prescription is what makes protocols useless.
- Margin and buffer widths
- With the scheme rates they attractBecause the environmental case and the financial case are the same case.
- Input thresholds
- When to spray, and when not toIncluding the flag leaf decision, which is the expensive one.
- What to measure
- Eight numbers, and how to take themEstablishment, tillers, ears, yield, blackgrass count, worm count, soil organic matter, and the gross margin.
How it changes
- A version a year
- At mostA protocol that changes every season cannot be evaluated.
- Changes come from returns
- Not from opinionA change needs data from adopting farms, and the number of farms is published alongside it.
- Nothing is removed
- Old versions stay publishedA farm that adopted v1.1 needs to be able to read v1.1.
- Failures are versioned too
- And publishedv1.2 recommended an earlier drilling window and it was wrong. That is in the changelog, permanently.
Why open rather than certified.
Because a certification scheme has to be policed, and policing costs money, which means membership fees, which means the scheme’s income depends on the number of members rather than on whether the method works. Every farm assurance scheme in Britain has this problem and most of them have it badly.
An open protocol has no members to lose. It is a document. Its only asset is whether the farms that follow it get better results, and the only way to know that is if they report back, which they will only do if there is nothing to lose by reporting a failure.
That is the whole design: no fee, no badge, no audit, no removal from a list. Just a version number, a set of numbers to measure, and a published table of what happened on the farms that tried it — including the ones where it did not work.
What adopting farms would report, and what happens to it.
| Measurement | How often | Published as |
|---|---|---|
| Establishment, plants per square metre | Once, at emergence | A distribution across all reporting farms, never a named figure |
| Blackgrass count, heads per square metre | Once, in June | The headline number, because it is the one the method exists for |
| Yield, tonnes per hectare | At harvest | A distribution, with soil type and rainfall alongside |
| Gross margin, per hectare | After harvest | Voluntary, and the most useful of the eight |
| Soil organic matter | Every third year | Because annual testing measures the laboratory rather than the soil |
| Worm count | Once, in spring | Cheap, unfashionable and a genuinely good proxy |
| Rainfall | Monthly | Without which none of the above means anything |
| What was ignored | Always | Which parts of the protocol the farm chose not to follow, and why |
How a farm would adopt it.
Read the current version
It is a document on a website. There is nothing to join and nobody to tell.
Adopt as much of it as makes sense
Partial adoption is expected and is explicitly reported rather than treated as failure.
Measure the eight numbers
Which takes about four hours across a season, most of it in June.
Send them back, or do not
Voluntary. A farm that adopts the method and reports nothing has still had the method.
Read what everybody else found
Published annually, as distributions rather than as named farms.
The reason this is worth doing is that British agronomy advice is overwhelmingly delivered by people who also sell inputs. That is not corruption and it is not a conspiracy; it is simply the structure of the industry, and it means a genuinely independent, openly versioned method with published results has no obvious competitor.
There is no protocol, and no farm has adopted anything
This page describes what an open agronomy protocol would look like. This is a demonstration build with no method published, no version number issued and no returns from anybody.
The method, questioned.
Because there is nothing to lose. No membership, no badge, no audit and no list to be removed from. The only reason schemes do not get honest returns is that honesty costs the member something.
Advice is verbal, unversioned and unaccountable. A protocol has a version number, a changelog and a published record of what happened when farms followed it, including the version that was wrong.
Because combinable crops on thin calcareous soils is what this farm actually is. A method claiming to work on every soil in Britain would be a marketing document.
And the research partner who would want the data.
A university, a working farm, and a fair exchange.
