On-farm trials
On-farm trials that give a straight answer at harvest
Lay strips or plots on your own field at implement width, send them to the planter or spreader with the prescription, and compare the treatments on cleaned harvest yield and the application rate that actually went on.
✓Strips and plots at your machine width
✓Trial and zones in one file for the display
✓Each treatment compared on the same ground
Independent of input sellers, and your trial data stays yours
30-day trial on up to 250 acres (101 ha), no card required
Trial designer
How many strips does a real answer take?
On uneven ground, one pair of strips measures the soil under them more than the seeding or nitrogen rate. Replicated strips side by side cancel the ground out, and the margin around the result shrinks with every block you add.
Crop
Treatment
Treatments in the trial
Between neighbouring seeding ratesBetween neighbouring nitrogen rates
Field
Strip width (implement)
The trial on the field
6 strips · 10.9 ac (10 % of the field)
30,000 seeds/ac
Your current seeding rate,Your current nitrogen rate, 34,000 seeds/ac
38,000 seeds/ac
Background: the field’s productivity zones
The answer
Each treatment against the one you use now, block by block
38,000 seeds/ac against your current seeding rateyour current nitrogen rate
+5.1 bu/ac give or take 4.0
The difference is real
Too close to call
One pair of strips gives no margin at all
30,000 seeds/ac against your current seeding rateyour current nitrogen rate
−5.3 bu/ac give or take 4.1
The difference is real
Too close to call
One pair of strips gives no margin at all
Same trial, the next season’s weather and soil noise
Every strip at harvest
bu/ac
Out of 100 seasons like this one
By number of blocks, 300 simulated harvests each
With 2 blocks, the trial finds the real difference in 10 of 100 seasons and points the wrong way in 4. From 8 blocks it finds it in 80 or more.Within 8 blocks it never reaches 80 of 100: a larger difference, or more even ground, needs to be the question.
Strips in each block sit side by side in a random order, so the ground under them is shared and the seeding or nitrogen rate is what differs. The margin is the 95 % range of the block-by-block difference. Strip-to-strip noise left after blocking: 2 % of crop yield on even ground, 3 % on typical ground, 5 % on patchy ground.
From a winter wheat trial
The higher nitrogen rate won by 1.5 t/ha, until the trial was read on the same ground
A field at 150 kg of nitrogen per hectare, with two trial areas at 120 and 180. The raw averages pointed one way; the same harvest read on the same ground pointed another.
Crop yield, t/ha
Both trial areas harvest more than the rest of the field by 1.4 to 1.5 t/ha, the lowest nitrogen rate as much as the highest. The nitrogen rate is not what differs here.
Keeping only the ground where the spreader held its nitrogen rate changes little: the trial areas still sit far above the rest of the field.
On the same high-yield zone the three nitrogen rates harvest 6.33, 6.34 and 6.41 t/ha. The trials had been placed on the field’s best ground.
On that zone, 60 kg/ha of nitrogen between the lowest and highest nitrogen rate bought no measurable gain in crop yield.
›The full analysis, step by stepTrial layouts
Five layouts, one for each kind of question
Width follows the implement, the angle follows the AB line, and a buffer keeps neighbouring treatments from mixing. Shuffling puts each treatment in a random place inside every block.
Field strip
Full passes across the whole field, crossing every zone.
Best for seeding or nitrogen rates with farm machinery, 4 to 6 replications.
Zone strip
Strips clipped to each management zone, so every zone gets its own comparison.
Best for asking whether the seeding or nitrogen rate should differ by zone.
Plot grid
Blocks of short plots, each block holding every treatment once in a random order.
Best for hybrids, varieties or products on fairly even ground, 3 to 6 blocks.
Split-plot
Seeding or nitrogen rates as the main plots, a second factor such as the hybrid repeated inside each one.
Best for the same seeding rates across two or more hybrids.
Checkerboard
Treatments alternating across a grid, spread over the whole trial area.
Best for a first look at a field with strong variability.
In GeoPard
From the zones map to the verdict, on one field record
Before the pass
Start from the zones map
Trials are drawn on the zones map you already use for the prescription, so the trial and the prescription for the rest of the field live in one layer.
›Tutorial: zones mapsLay out the trial
Pick the layout, set width to the implement and the angle to the AB line, add a buffer, and set the product rate of each treatment around the one you use now with one slider. Shuffle, check the fit on the map, save.
›Tutorial: trial design
In the season
Send it to the cab
The trial exports merged with the zones as one prescription: ISO-XML for an ISOBUS terminal, or a John Deere work plan straight to Operations Center.
›Tutorial: John Deere work plansCheck what went on
The as-applied map against the target application rate shows which strips got their application rate and which did not, before anyone reads the harvest.
›Tutorial: applied against target
After harvest
Clean the harvest yield data
Calibrate combines that ran side by side and drop turnarounds, stops and spikes, so a strip’s harvest yield reflects the crop and not the machine.
›Tutorial: yield data cleaningRead the result
Compare treatments on the application rate that went on and inside the same productivity zone, then see what each product rate returned in the Yield Report.
›Tutorial: trial analysis
Who runs trials on GeoPard
A new seeding or nitrogen rate proved on one field before it goes on every field
Growers
Test a lower seeding or nitrogen rate on your own ground with your own planter, and change the whole farm only when the strips say so.
Agronomists and dealers
Run one protocol on many clients’ fields, compare hybrids and seeding rates in split-plots, and hand each client a result read on the same ground.
Researchers
Every plot keeps its position on the field, next to harvest yield, soil, elevation and as-applied layers, ready for the statistics you already use.
“GeoPard’s analytics have made it easy for our team and farmers to use precision ag tools effectively. Reliable analytics and consistent support throughout the season have improved input planning and management.”
Edoardo PozzatoPrecision Agriculture AgronomistSergio Bassan, ItalyThe dealer’s trial workflow“It is a pleasure for us to work with the enthusiastic GeoPard team. Deeper insights into our strip-intercropping data enabled by the GeoPard tool are very valuable to us.”
Pricing
Trials are included
A trial is drawn on the zones map and exports with the prescription, at no extra charge. Annual covers every field; on Pay-As-You-Go you pay only for the zones or prescription map the trial sits on.
Annual
The whole farm, every trial and every field, all season.
$0.59 per acreper hectareper year on the first band, down to $0.45 as your area grows
- ✓Unlimited zones, prescriptions and equation maps
- ✓Minimum $200 a year, covering up to 339 acreshectares
- ✓One user seat included, extra seats $60 a year
Pay-As-You-Go
A trial on a few fields, or a first season.
$20 a monthplus each operation you run
- ✓Field registration: $0.10 per acreper hectare, once
- ✓Zones or VRA map, with the trial: $0.072 per acreper hectare
- ✓Yield data cleaning: $0.024 per acreper hectare
- ✓Equation map: $0.10 per acreper hectare
30-day trial
Your own fields, before you pick a plan.
250 acres (101 ha)for 30 days, no card required
- ✓Every analysis feature, trial layouts included
- ✓John Deere Operations Center connection
- ✓Nothing is charged when it ends, and your fields stay in your account
Billed in USD in acre regions such as the US and Canada, and in EUR everywhere else.