One of our daughters had a weird dream over Christmas. Apparently, I went missing for six months and was eventually found on a golf course. When asked why I had not returned home earlier I simply replied that I was having a bad round. Obviously, at least one of our daughters thinks that I have fixations. Golf is probably one and another is definitely looking at how data can provide guidance to farmers.
I have recently been looking at datasets of nitrogen trials on winter wheat. This is a risky exercise because of the variable nature of results and the way optimum doses are calculated. Imposing any assumption on the shape of the response can influence the results and maybe mask the true answer. However, using the conventional methods, which may not be correct for some individual trials, the datasets I have examined allow me to have a more informed deduction as to how to get high wheat yields with nitrogen. It seems that the way to get high nitrogen-fertilised yields is to get high yields in the absence of applied nitrogen. In general, the higher the unfertilised yields, the higher the fertilised yields. I think that most farmers either know or suspect this to be the case. Good crop and soil management and favourable weather conditions are the key, with nitrogen providing the icing on the cake.
In these databases, for every extra tonne of nitrogen-fertilised yield/ha there is a contribution of at least half a tonne from the yield without applied nitrogen. One very large database suggests that for every tonne of nitrogen-fertilised grain/ha, two thirds of a tonne comes from the yield without nitrogen. These observations have a surprisingly high level of statistical validity for nitrogen trials. The associations explain to me why the additional amount of nitrogen required for very high yields in field trials is less than would logically be expected.
I have to say that the contribution of yields without nitrogen to increasing fertilised yields vindicates the nitrogen recommendations for feed wheat, based on field trials, provided to NIAB TAG members over the last ten years or so. These recommendations ignore Soil Mineral Nitrogen (SMN) levels if below 100 kg N/ha (i.e. nitrogen indices 0-3 in the current version of RB209) and only suggest applying relatively small amounts of additional nitrogen to a realistic base dressing when field yields are expected to exceed 10-11 t/ha. This is in contrast to the current version of RB 209 where the recommended doses decrease significantly between nitrogen indices 0-3. The reason why SMN can normally be ignored at these levels of SMN is because potential yields tend to increase with more available soil nitrogen: therefore the crop can economically use it and the additional nitrogen applied over and above the reduced doses recommended in current RB 209 for nitrogen indices 2-3. This is a rather neat self-correcting system. It appears to explain why there is, on average, no real difference in the economic optimum nitrogen dose in field trials carried out in situations where SMN is anywhere below or around 100 kg N/ha.
The conclusion of this piece is to reinforce my usual plea that we should be more cognisant of what field trials are telling us, despite their associated frustrations and errors. Now that I have got that off my chest I must return to the golf course.
Best wishes for 2017 (and don’t overdo the nitrogen).