Winter oilseed rape: Effects of sulphur on seed glucosinolate content and seed yield

Abstract

To determine the effect of sulphur (S) fertilisers on seed yield and glucosinolate (GLS) content of double low varieties of winter oilseed rape, ammonium sulphate (10, 20, 30, 50 and 80 kg S/ha), gypsum (20 and 50 kg S/ha) and foliar elemental sulphur (20 kg S/ha) were applied in spring to 4 crops on shallow chalk soils and one crop on a sandy soil in areas where the rate of annual deposition of sulphur was less than 20 kg/ha.

Application of S significantly increased yield at the site on sandy soil which was in North Northumberland and which showed severe S deficiency symptoms from stem extension onwards. There was also a significant yield response to increasing rates of ammonium sulphate at a shallow chalk site in Dorset which did not show symptoms of S deficiency. There was no significant advantage from putting on more than 10 and 50 kg S/ha respectively at these sites. At these rates of applied S respective yield responses of 15 and 12% were best predicted by the concentrations of total S in young fully expanded leaves at flowering which were at or below 0.36%.

Increasing rates of ammonium sulphate gave small significant increases in seed GLS at 3 of the sites on chalk soils but at the site on sandy soil seed GLS was trebled at rates of 50 and 80 kg S/ha applied. Application of S decreased oil content at sites on chalk soils by an average 0.3% but at the site on sandy soil the reduction was over 1% and significant. When compared at a single rate of 20 kg/ha, the effectiveness of the different forms of S applied in raising plant S status was in the order ammonium sulphate > gypsum > foliar elemental sulphur although differences did not significantly affect yield response.

The pattern of yield and GLS response in North Northumberland has been found in previous experiments and confirms that sandy soils in this locality are very susceptible to S deficiency. Shallow chalk soils in Dorset have also previously shown yield response to applied S. The results reported here suggest that further research is required to more accurately quantify the supply of S from the soil.