It is wishful thinking to believe nitrogen inputs to cereals can be spot-on every time. But too few growers use the tools that can help them get closer to optimum dressings, said ADAS’s Roger Sylvester-Bradley in a presentation to AICC members.
Highlighting the uncertainties determining the amount of N available to crops, not least its release from soil organic matter, Mr Sylvester-Bradley said: “We have to accept that we’ll never get it right every time.”
Experience from 47 recent HGCA trials showed, with hindsight, how difficult it was to hit the optimum input target, he said. Only half the applications came reasonably close, and in nine cases the recommended amounts were 100kg/ha over or under what turned out to be the optimum.
“The average lost profit was 0.15t/ha of grain,” he said. “It’s not a huge amount, but the work does suggest an appalling level of uncertainty.”
Growers’ best strategy in such circumstances was to concentrate on eliminating applications furthest from the optimum, he advised.
“Forget about fine-tuning – make sure you get the outliers right.”
Diseases and weather during the growing season clearly made it hard to forecast crops’ nitrogen needs, Mr Sylvester-Bradley acknowledged. “Our ability to predict yield and therefore N demand is not very good.”
But knowing what was potentially available from the soil was a vital starting point.
Assessing top-soil organic matter levels was a useful, cheap, initial screen, but testing for soil mineral nitrogen (SMN) to 90cm was the best option, he said. “People tend to wait too long to do it. It can be done from November right through to March, and it rarely underestimates crop-available N.”
Errors in applications accumulated over seasons, he warned. Up to 40% of excessive dressings could build up within soil organic matter, so monitoring was crucial.
However, the top of the yield-response curve to N was “reasonably flat”, so an extra 10kg/ha could be seen as good insurance, he said.
Grain nitrogen analysis remained a useful retrospective check on the impact of particular N dressing tactics, and it was worth noting that in 60 HGCA trials, 20% of fields needed no N fertiliser at all. “There are fields on livestock farms that certainly don’t need N fertiliser,” he added.
Providing a reliable indication of N supply from soil remained a key challenge, said Keith Goulding of Rothamsted Research.
“SMN testing is useful, but there are still problems in interpreting the results.”
At low values, the system tended to underestimate the N supply and vice-versa at high values, he said.
“We’re trying to get DEFRA funding to look into this.”