Abstract
Alfalfa (Medicago sativa), rape (Brassica napus) and barley (Hordeum vulgare) were grown in a growth chamber with and without sulphur (S) on surface samples (0-15 cm) of 17 eastern Canadian prairie soils. The soil varied in pedological origin and were divided into three groups based on their total nitrogen:total sulphur (N:S) ratios. Herbage yield of alfalfa (full bud) and rape (flowering ) were highly correlated (P < 0.01) with soil (N:S)s ratios and soil SO4-S, while for barley (shot blade) the relationships were significant (P .ltoreq. 0.05). Herbage yields of the three crops were also significantly correlated (P .ltoreq. 0.05) with plant (N:S)p ratios. Soil (N:S) ratio was considered the best predictor of herbage yield because it is more stable over time than soil SO4-S and because plant (N:S)p ratio only indicates a need for future fertilization. The highest yields and S and N composition of plant tissues were obtained on soils with SO4-S concentrations greater than 12 mg kg-1 and an (N:S)s ratio of approximately 6. These soils had sufficient S to meet the requirements of all three crops. Addition of S had no effect on herbage yield but did increase S uptake. Results indicated that soils with (N:S)s ratio of 9 and SO4-S concentration of 7 mg kg-1, may have sufficient S to meet the requirement of barley but not alfalfa and rape. However, soils with higher (N:S)s ratios and lower SO4-S content were deficient in S for all crops. Barley was the most responsive to applied S of the three crops, although it had the lowest concentration of S (0.15 mg S g-1 dry herbage)and highest plant (N:S)p ratio (16) at its highest yield. Rape was more responsive to applied S than alfalfa but the concentration of S in the herbage of both crops at their highest yields were similar, 0.27 and 0.26 mg S g-1 dry herbage, respectively. Both crops also had lower (N:S)p ratios than barley at the highest yield (14 (alfalfa) and 12 (rape)), reflecting their greater S uptake.