Variable-Rate Seeding: When It Pays in UK Conditions

Most variable rate seeding sales pitches promise to cut your seed bill. On a lot of UK fields that doesn’t happen. A good prescription usually moves seed from one part of the field to another, and the total used can stay the same or even rise. The money comes from two places: fields where establishment differs a lot from zone to zone, and drilling dates late enough that seed rates are already high. If neither applies to a field, a flat rate will serve you just as well.

The rest of this piece works through the numbers on one example field. It then sets out a strip trial you can run this autumn to test the return on your own ground.

The one formula behind every seed rate

Every seeding prescription, whether it comes from SOYL, Hutchinsons Omnia, Agrii or John Deere Operations Center, uses the same sum that AHDB’s Wheat Growth Guide sets out:

Seed rate (kg/ha) = target plants/m² × thousand grain weight (g) ÷ expected establishment (%)

Take a target of 250 plants/m², a TGW of 50 g and 70% establishment. That gives 250 × 50 ÷ 70 = 178.6 kg/ha.

Target plants and TGW are the same across a field. Establishment isn’t. Seed on a sandy loam knoll with a fine tilth might come through at 80%. On a cloddy, wet clay hollow in the same field it might manage 45%. A “variable rate” prescription is really a variable establishment figure fed into the formula. If your zones all establish at roughly the same rate, the formula gives roughly the same seed rate everywhere and there’s nothing to gain.

A worked example: 24 ha of winter wheat, two drilling dates

The field is a 24 ha block typical of the Midlands or East Anglian clay margins:

• 6 ha of sandy loam on the rise
• 10 ha of medium loam across the middle
• 8 ha of heavy clay in the bottom

Seed is certified C2 at £550/t (£0.55/kg), with a TGW of 50 g.

The establishment figures below are working assumptions in line with the AHDB ranges. The pattern is the important part: the gap between zones widens as the season gets later. Clay drills well in a dry late September. By November it’s cold and sticky, and slugs are active.

Most growers on a field like this set a flat rate that protects the worst zone, because nobody wants a thin crop on the clay. That flat rate is what variable-rate seeding has to beat.

Drilled 25 September, target 250 plants/m²

The prescription saves 338 kg, which is £186 across the field or £7.75/ha. Soil scanning and prescription maps cost around £8/ha a year once you spread a one-off EC scan (often quoted at £15 to £25/ha) over five seasons and add an annual mapping fee. Early drilling on this field breaks even at best.

Drilled 5 November, target 300 plants/m²

This time the prescription saves 1,544 kg. That’s £849 across the field, or £35.40/ha. Take off the £8 mapping cost and you’re left with about £27/ha. If you use farm-saved seed at roughly £0.35/kg after cleaning, dressing and royalty, the saving is still £22.50/ha before mapping.

The flat rate also has a cost that the seed bill doesn’t show. At 667 seeds/m² and 70% establishment, the sandy knoll ends up with about 467 plants/m². That’s a crop heading for lodging, more PGR spend and a messy harvest on the bit of the field that should be easiest.

Growers who drill at the field average instead of the clay rate face the opposite problem. A flat 500 seeds/m² in November leaves the clay at 225 plants/m² against a 300 target. Late-drilled wheat has little time to tiller and make up for missing plants, so that zone can lose yield. In that case variable-rate seeding uses more seed overall, and the payback has to come from the clay yield.

Either way, the conclusion is the same. Variable-rate seeding pays on fields where establishment varies widely, and the later you drill, the wider that variation gets. A uniform field drilled in September gives you almost nothing back.

Where the “AI” actually sits

Most of the intelligence in these products goes into drawing the zones, not working out the rate. Omnia, SOYL’s SOYLsense, Agrii’s iSOYLscan-based services and Climate FieldView all combine soil EC scans, several years of yield maps and Sentinel-2 satellite imagery, then group areas of the field that behave alike. That grouping, a form of clustering, is the part that’s hard to do by eye. For more on the data layers behind those maps, see our guide to satellite, drone and variable-rate agronomy.

The software doesn’t know your establishment figures. Most platforms come with default establishment assumptions, and you should change them. If you know a field’s clay comes through at 50% in a wet November, enter 50%. Leaving a rosy default in place is the most common reason a VR map does no better than a flat rate.

On the drill, any ISOBUS machine with electric metering and a Task Controller licence will take a prescription file (ISO-XML or shapefile). That includes Horsch, Väderstad (through E-Control or a Seed Hawk set-up), Amazone, Kuhn and Sumo machines. Check that your terminal has rate control unlocked before you pay for maps. Plenty of farms are running drills that could do this and a terminal licence that can’t.

Run your own strip trial this autumn

Don’t take a supplier’s word for any of this. A strip trial costs you an hour of setup and some plant counting in February. It also gives you a figure that applies to your soil, your drill and your slug pressure.

1. Choose the right field. Pick one with at least three soil zones you already know about: the field where you’d hesitate before setting the rate. Leave out your most uniform block, because it will just show no difference and teach you nothing.

2. Lay out alternating strips. Mark at least three VR strips and three flat-rate strips side by side across the zones. Each strip should be one drill width or a multiple of it, and also line up with your combine header so each harvest pass stays inside one strip. Save the boundaries as a guidance pattern in your terminal so you can find them again at harvest.

3. Set the flat rate to your normal practice. If you usually drill at the clay rate, use that. Comparing VR against a flat rate you’d never use proves nothing.

4. Count plants in February. In each zone of each strip, take five counts along 1 m of drill row. Divide by the row spacing in metres. For example, 30 plants in 1 m of 12.5 cm rows is 30 ÷ 0.125 = 240 plants/m². Note the figures on your phone against the strip number.

5. Clip the yield map at harvest. Export the yield data and cut it by strip and by zone. Drop the headlands and the first 20 m of each zone boundary, where the metering is still adjusting. If you don’t have a yield monitor, weigh each strip’s grain over a weighbridge or with trailer load cells.

6. Work out the margin per zone. Value seed saved at your seed price and yield gained at a realistic grain price. Here’s an example of how the summary might look, using illustrative figures with wheat at £180/t:

Zone    Strip  Seed kg/ha  Plants/m²  Yield t/ha
Sand    VR     214         301        8.9
Sand    Flat   333         452        8.7
Loam    VR     250         298        9.4
Loam    Flat   333         395        9.4
Clay    VR     333         302        8.1
Clay    Flat   333         297        8.2

Margin per ha, VR vs flat (seed £0.55/kg, wheat £180/t)
Sand:  119 kg x 0.55 = £65.45 seed  + 0.2 t x 180 = £36.00   -> £101.45
Loam:   83 kg x 0.55 = £45.65 seed  + 0.0 t          = £0.00  -> £45.65
Clay:    0 kg                         - 0.1 t x 180 = -£18.00 -> -£18.00

Field-weighted: (6 x 101.45) + (10 x 45.65) + (8 x -18.00) = £921 / 24 ha = £38.40/ha
Less mapping £8/ha                                           = £30.40/ha

Look at the clay row. Both strips got the same seed rate, so the 0.1 t/ha gap comes from soil variation and combine error, not from the treatment. That figure is your noise floor: in this field, any yield difference smaller than about 0.1 t/ha means nothing. Results like the extra 0.2 t/ha on the sand only count if they come up in most of your replicate strips, not just one.

What to do with the result

If the trial shows £20/ha or more in a late-drilled, variable field, pay for zone maps on your other fields with similar soils and drilling dates, and keep flat rates on the uniform ones. If it shows nothing, you’ve found that out for one field’s worth of seed and a few hours of counting rather than a year’s mapping contract.

Next autumn, swap which strips get the VR rate and drill on a different date if the weather lets you. Two seasons of results from your own field will tell you more than any figure a supplier can quote.