Optimisation de la fumure soufree par estimation du risque de carence. Ⅱ. Ble d'automne

摘要

Results of field experiments conducted during four growing seasons were used to improve a model for the risk assessment of sulphur deficiency in winter wheat (RYSER et al., 2001). The scoring model based on agronomic, pedologic and climatic criteria explained 84% of yield variability in response to sulphur application. In order of importance, significant parameters of the model were soil stone content, rainfall between October to March, soil organic matter and plant-available nitrogen. Moreover, clay content, soil depth and application frequency of organic fertilisers were also taken into account. The risk of deficiency was divided into 3 categories and sulphur recommendations were made for all of them. The low risk of deficiency was associated with a score of more than 18 points and no S fertilization was required for this category. The optimum S fertilisation of the moderate deficiency risk with a score of 13 to 18 points was set to 10 kg S/ha. The high risk (score under 13 points) was not represented in the experiments but a recommendation of 20 kg S/ha has been made. Sulphur application increased wheat yield significantly (by 15% on average) for moderate risk sites but had no effect on low risk ones. Total sulphur uptake in the aerial biomass was increased by sulphur application. Breadmaking quality was also evaluated. The total protein content of the grains was not affected by sulphur application. By contrast, 3 out of 4 moderate risk sites and only 2 out of 9 low risk sites showed significant increase of Zeleny values. This scoring model for the risk assessment of sulphur deficiency should allow farmers to find the optimal sulphur fertilisation in winter wheat.