Digital N rate advice hope
Computer modelling to
pinpoint best crop nitrogen
dressings still has a way to
go before growers can use it
confidently. Andrew Blake
reports from a recent
meeting of researchers, crop
consultants and farmers at
FIRST results from on-farm tests of SUNDIAL-FRS from Yorks to Kent last year confirm that optimum N application ranges widely according to site (table 1).
That more than justifies attempts to refine recommendations, suggested Rothamsteds Liz Stockdale. "Clearly for each crop type there is no specific optimum. Site specificity is all."
SUNDIALs best estimate of the needs in a year clearly wetter than normal was closer to the optimum than standard MAFF booklet RB209-based calculations in two of the three winter wheats after winter oilseed rape last year (table 2).
But for the third it would have been better to have stayed with the conventional approach.
Other N-need indicators tried by some farmers at the meeting included Hydros PrecisioN Plan, Kemiras N-Min system, the French Jubil approach and use of Solomons porous cups.
SUNDIAL, a LINK project funded by MAFF, HGCA, BPC, HDC and SBREC, was tried on replicated small plots at 18 farms with the help of AICC members. All inputs bar nitrogen remained the same across the fields. For each site five N rates were applied, two above and two below a mid-dose guided by the computer program.
There were also zero N plots at each site to help determine overall response curves.
Yields ranged quite widely. But even without added N fertiliser one site delivered 6t/ha (2.4t/acre) of wheat, highlighting the value of soil reserves, she noted.
First-year trials at the Scott Abbott Arable Crops Station comparing six different assessment methods, including the sap test-based Jubil, were outlined by Matt Spence.
Brown rust restricted potential yield of the Riband wheat, he acknowledged.
With wheat at £70/t and Nitram £100/t, the economic optimum yield was 9.3t/ha (3.8t/acre) using 72.5kg/ha (58 units/acre) of N. That was only just below maximum output on the thin limestone soil which came using 76.6kg/ha (61 units/acre), Mr Spence noted.
The prediction closest to the optimum came from Kemiras N-MIN which initially advised applying 57kg/ha (46 units/acre). But that was before the soils shallowness had been taken into account, he said. "When we did that we found the rate should have been 107kg!"
No other method came near actual optimum. Furthest out was RB209. Assuming an N index of 1 after the peas it advised 190kg/ha (152 units/acre). Hydro, SUNDIAL, Jubil and N-Able (a veg-orientated model developed by HRI) recommended 120, 140, 150 and 160kg/ha (96, 112, 120 and 128 units/acre) respectively.
Jubils shortcoming was its reliance on applying most of the N early, leading to excessive tillering which could not be supported later on the thin soil, said Mr Spence. Final yield was 7.5t/ha (3t/acre). The same amount applied as a three-way split delivered 8t/ha (3.3t/acre).
• One way to fine tune SUNDIAL during the season could be to incorporate field measurements of the pool of potentially available nitrogen in the soil, for example as indicated by the Solomon system. *
Table 2: N advice comparisons
Optimum & recommended N rates on three sites (kg/ha)
Optimum SUNDIAL RB209
Site A 90 95 175
Site B 116 125 220
Site C 177 130 190
Table 1: Best N
Economic* optimum N fert ranges (kg/ha)
First wheat after legumes 0-155
First wheat after OSR 76-177
Second wheat 108-170
Winter barley 115-122
* Wheat @ £60/t, N fert £100/t.