The effects of site, season and sulphur and nitrogen fertiliser on yield and seed glucosinolates of winter oilseed rape

Abstract

The early stages of this project examined the factors causing, and the processes underlying, site and season variation in seed glucosinolate concentrations in winter-sown oilseed rape. The later stages concentrated on the yield responses to applied sulphur and the possibility of predicting these by soil or crop analysis. The project involved a series of experiments conducted between 1990 and 1993 at five sites known to differ in their sulphur status. The experiments tested interactions between site, season and combinations of nitrogen and sulphur treatments, sometimes on a range of cultivars. Measurements were made of crop growth, nitrogen and sulphur uptake, leaf sulphur concentrations, and patterns of growth of seed and their accumulation of nitrogen, sulphur and glucosinolates.

Seed yields ranged from 0.4 to 5.0 t/ha. They were increased by nitrogen at all sites in all three years. There were small responses (<14%) to applied sulphur at Cockle Park in 1991 and 1992, in the Scottish borders in 1992, at ADAS Bridgets in 1992 and 1993, and at Rothamsted in 1993. Applied sulphur doubled the yield of a small crop at Woburn in 1991 and increased yield by 37% in 1992.

Seed glucosinolate concentrations in the cv. Falcon ranged from 8 to 31, µmol/g and varied with site and season. Applied sulphur had no effect on concentrations at Rothamsted and Cockle Park, but greatly increased them in the Scottish borders and at Woburn and ADAS Bridgets. The response to applied nitrogen depended strongly on the sulphur status of the crop. Seed glucosinolate concentrations were increased by applied nitrogen when the sulphur status of the crop was high, and decreased when the sulphur status was low. Leaf sulphur concentrations were not a reliable predictor of the responses of yield and seed glucosinolates to applied sulphur.

The variation in seed glucosinolate concentration between sites was not related to the ratio of crop sulphur to seed number/m2. Glucosinolate concentrations were related more strongly to seed number alone.

Seed glucosinolate concentrations were measured by both the X-RF and HPLC methods. There was evidence that the X-RF method, even when corrected for seed nitrogen content, could give unreliable values for young seed or seed from sites where the crop nitrogen and sulphur are widely unbalanced.