How do I test manure or compost for weedkiller before I use it?

There is no home chemical test for aminopyralid, and laboratory testing costs more than the manure is worth. What works instead is a plant so sensitive to these herbicides that it reports concentrations no instrument in a garden could find. Beans are that plant, and the test costs the price of a packet of seed.

Safety

What you'll need

• Three or more pots of the same size, at least 1 litre
• Broad bean or French bean seed, nine or more
• Bagged compost from a source you already trust
• The suspect manure, compost, straw or hay
• Labels and a pen
• A warm, light place — windowsill, greenhouse or propagator

Step by step

1. Take a representative sample of the suspect material.Not one handful from the top. Take from four or five points through the heap, including the middle and the base, and mix them. Contamination is not evenly distributed through a stack, and a sample from one spot can miss it.
2. Set up the control pot.Fill one pot with plain multipurpose compost from a bag you have used successfully before. This is what everything else is compared against, and without it the test tells you nothing — beans grow oddly for all sorts of reasons and you need a baseline.
3. Set up two test pots at different strengths.One with half suspect material and half trusted compost, mixed thoroughly. One with the suspect material neat. Two strengths catch mild contamination that a single dilution might miss, and they show whether the effect scales with dose, which is what confirms the diagnosis.
4. Sow three bean seeds in each pot.Broad beans or French beans, sown to their usual depth. Three per pot so that one failing to germinate does not cost you the test. Peas work too but are slightly less sensitive.
5. Label every pot clearly and immediately.Control, test A half, test B neat. In three weeks you will not remember, and an unlabelled bioassay is worthless.
6. Grow them side by side in identical conditions.Same windowsill, same watering, same temperature, same light. Every difference you introduce is a difference you might misread as damage. Somewhere warm matters — cold slows growth to the point where the test takes six weeks to read.
7. Wait three weeks, and resist reading it early.The first true leaves are the ones that report. Mild contamination often looks normal for the first ten days and distorts as the plant grows. Two weeks is not long enough on a mild case.
8. Read the result against the control.Clean: all three pots have straight stems, flat even leaves, and the same colour and size. Contaminated: the control is normal while the test pots show cupped leaves curling down at the edges, narrowed fern-like leaflets, twisted or fused growing tips, and stunting — usually worse in the neat pot than the half-and-half pot.
9. Act on a positive result.Do not use the material anywhere you grow susceptible plants. It can be spread on lawns, on paths, around established woody plants and around sweetcorn, onions, leeks and brassicas, where it does no harm and will break down over time. Tell whoever supplied it.
10. Repeat if the answer is ambiguous.Cold, slugs, overwatering, poor light and compacted compost all produce mild distortion that can be mistaken for a positive. If you are not sure, rerun with fresh seed and better conditions before condemning a load.

Tips

• Run the test in advance of when you want the material. Three weeks in February is easy; three weeks in the middle of planting is not, and the temptation to spread it untested is exactly how gardens get contaminated.
• Test each delivery separately. A yard that was clean last year may have bought different hay this year, and the risk does not carry over.
• The same test works on straw for mulching, hay, home-made compost you have doubts about, and bulk green-waste compost. Anything that can carry the residue can be tested this way.
• Keep the pots after a positive result and photograph them alongside the control. If you need to raise it with a supplier, a photograph of the two side by side is far more persuasive than a description.
• A shorter, rougher version — sowing beans directly into a patch of the suspect area outdoors — works in warm weather but is much harder to read because you have no matched control.

If it doesn't work