Growers have long conducted side-by-side tramline trials to help fine-tune their management, however, this unscientific approach has never been taken seriously by researchers.

That is until now, as new advances in precision farming technology, data recording and analysis could see a shift to more on-farm research providing practical advice to help drive up crop yields and cut costs.

See also: Farmer embraces data to improve crop management

For more than a century, scientists and research companies have relied on small replicated and randomised plot trials to carry out their work.

Although this valuable tool has many benefits, trials are expensive to run and only give results over a very small area.

This makes it hard to judge the relevance of trial findings across regions, farms or even within fields due to variability in soils and weather.

On top of this, assessing the impact of practices such as cultivations, rotations and straw incorporation is difficult to factor in to plot trials.

This is because it is tricky to set up treatments in a way that accurately mimics commercial practice and effects need to be assessed over a number of years.

Lack of precision can also be a problem, as conventional trials are typically only able to confidently detect wheat yield differences of about 0.5t/ha.

However, Daniel Kindred, senior research scientist at the Agricultural Development and Advisory Service (Adas), notes that many management decisions cost under £30/ha.

This equates to a yield benefit of less than 0.3t/ha at current wheat prices.

“Increasing yield is about these small incremental gains, so we need to be able to detect them in trials at a scale that is relevant to growers and helps make decisions that are financially profitable for their businesses,” he explains.

New approach

With this in mind, Adas and its partners have developed a new approach to research called “Agronōmics”. It allows more accurate insights into treatment differences from larger scale on-farm trials carried out by growers using conventional machinery.

Like small plot trials, on-farm trials have their limitations too, as the number of treatments they can compare is limited and larger areas inherently have more variation.

However, the latest data-gathering technology can now take account of these variables.

Electromagnetic induction scanners and soil brightness imagers map soils accurately, while satellite imagery, aerial imagery from planes or drones or data gathered from tractor-mounted gadgets such as Yara’s N-Sensor can monitor crop performance in-season.

Finally, yield mapping can record results of any management changes and Dr Kindred believes data from on-farm trials can be as precise, if not more so, than conventional trials.

“Yield data can be noisy [contain errors] and does require some work, but we are coming up with a statistical model that will enable us to take out errors and accurately tell differences between treatments, even in situations with high spatial variability,” he explains.

Farm network

While the technology tackles variability in the field, the bigger picture comes from building a network of farms both locally and on a wider scale, taking part in similar trials to maximise benefits of the on-farm research for all involved.

It allows the amalgamation of data from multiple locations and on-farm experiments using IT systems, allowing “big data” from a range of situations over a number of years to better support crop management decisions.

Networking is already taking place in a number of initiatives, such as Adas’s own Yield Enhancement Network (Yen), AHDB’s LearN and Innovative Farmer “Field Labs”.

In the past, the data generated from such on-farm, practical research didn’t have a platform to be recorded, saved and shared like it has today, according to Vince Gillingham of precision farming experts AgSpace.

He says that past on-farm trials were typically out-of-a-can product comparisons and any differences seen were anecdotal, but now there is the possibility to back up results with coordination and hard stats.

“This is a game changer, as you can see effects from multiple crops, in multiple fields and across a range of different soil types and geographies.

“They can provide more value than sterile small plot trials and give farmers the confidence that results from such trials are actually relevant on farm,” says Mr Gillingham.

Case study: Jeremy Margesson, Wiltshire

Being part of the agronōmics project has provided one Wiltshire grower with valuable local data and reassurance that his current crop management is on the right track.

Jeremy Margesson of the Burderop Farming Company, which farms about 1,200ha of wheat, barley and oilseed rape near Swindon, was offered the opportunity to take part in some farm-scale tramline trials last year.

These looked at high, medium and low fungicide and nitrogen input programmes in his wheat with help from Adas and precision farming experts AgSpace, who provide the data platform and capture for the project.

Using the farm’s RTK (real time kinematic) guidance system, treatments were accurately applied according to application maps and the combine – equipped with yield mapping technology – recorded the results at harvest.

Although he is waiting for results to be verified, the high input programme looks to have provided an 8.5% yield benefit or about £5-£10/ha improvement on margin over the low input regime.

“We have seen that altering inputs was paid for by the extra yield. Rather than giving us information on what to change, it is giving us the comfort that what we are doing is right.

“However, it has only been one year and the weather is the biggest factor, so it is something you need to look at over a number of seasons,” explains Mr Margesson.

Better value

Mr Margesson adds that he has hosted traditional trials before, but aside from providing the field for a distributor or research body to establish its plots, involvement in the trial was minimal.

Conversely, last season’s farm-scale trial provided first-hand involvement in design and treatments and exclusive access to results relevant to his own farm.

As such, he is happy to continue his involvement as long as it doesn’t impact on the practicalities of his day-to-day operations and remains hassle free.

“It is difficult to take time out at key fungicide timings when dry spray days are at a premium – you can’t spend half a day on trials.

“It needs to be designed so it doesn’t affect workload and we have had some input into that to make sure it works out,” he says.