30 September 2013
31 October 2015

Towards predictive nitrogen fertilisation recommendations for field crops

Through its participation in the Interreg IV ‘SUN’ project (Sustainable Use of Nitrogen) CRA-W has played a part in giving Wallonia’s farmers a dynamic new tool, called AzoFert®, for predictive nitrogen fertilisation recommendations for field crops.

Towards predictive nitrogen fertilisation recommendations for field crops

Developed by INRA in Laon (département of Aisne, France) AzoFert® is based on the projected nitrogen balance at plot level. This is determined by balancing nitrogen demand and supply in order to predict the additional nitrogen fertilization needed for optimum crop development. The innovative feature of this tool, besides being based on a complete nitrogen balance, lies in its dynamic approach, that is to say, it takes weather conditions over the year into account when calculating the nitrogen supply and, in particular, the mineralisation kinetics of all organic sources considered. AzoFert® was adapted, improved and validated in Wallonia by CRA-W in the context of the SUN project over the period from January 2010 to June 2013.

The first step was to adapt the software to Wallonia’s pedoclimatic conditions. The values of several categories of parameters (crops, soils, farmyard manure) were accordingly altered to take account of the differences between northern France and Wallonia. The weather data needed to use the program were obtained through a cooperative link-up with the Pameseb association.

Following configuration, the software was then tested on data from 42 agricultural field trials conducted by CRA-W in different parts of Wallonia between 1995 and 2011 in cereal crops (winter wheat, winter barley, maize), beetroot, potatoes and vegetables (spinach, beans, carrots and escarole). The recommendations produced by AzoFert® were compared with those yielded by Azobil, the previous, simplified and static version of the AzoFert® program, which had been adapted and used by CRA-W in Wallonia since the early nineteen-nineties. The conclusions were that in nearly 80% of the trials the difference between the recommended rate and the optimum rate was less than 30 units (as against only 50% with the Azobil program) and that the nitrogen supplied by mineralisation of organic sources as predicted by AzoFert® was more accurate than as predicted by Azobil.

Being a modelling program, AzoFert® is thus not infallible; the calculated rate is a predicted rate based on a yield target or on fixed values for crops’ nitrogen needs and on weather data that are partly historic averages. The recommendations are therefore always surrounded by a degree of uncertainty, for instance during spells of extreme weather (drought, long periods of low temperatures, etc.) However, the recommendations can be refined during the growing period by using plant nitrogen status measuring tools combined with split applications of nitrogen fertilizers.  Likewise, after harvesting the residual nitrogen can be absorbed by sowing a catch crop.

Its dynamic approach qualifies AzoFert® as a benchmark tool, but before it can be generally used as such throughout Wallonia it will first have to be adapted by all the laboratories in the Walloon RequaSud network, as part of the current harmonisation of nitrogen fertilisation recommendations in Wallonia.

SUN is an Interreg IV project jointly funded by the ERDF and CRA-W, project FW 4.1.10:   http://www.sun-interreg.eu/

Useful information