Dry matter responses by component species of 18 established, commercial Macroptilium atropurpureum cv. Siratrolgrass pastures to gypsum topdressing treatments (0 or 25 kg S/ha) were assessed from field experiments conducted over a four-year period in south-eastern Oueensland under rain-grown conditions. The objective was to establish diagnostic criteria for the assessment of sulfur status by relating pasture yields to agronomic attributes and soil and plant chemical tests. Beneficial responses to gypsum were small (maximum of 32% in Siratro) and restricted to fewer than 25% of sites, whereas at disadvantaged sites (28% of total), grass yields were more severely depressed than Siratro yields. It was not possible to predict these effects from past sulfur fertilizer history, Siratro percentage in the pasture, or pasture age. Significant correlations between Siratro relative yields (100 x yield without sulfur/yield with sulfur applied, attenuated at 100 for model fitting) and both soil sulfate and plant sulfur concentrations confirmed the predictive value of these laboratory data. For Siratro, best prediction of responsiveness was provided by sulfur concentrations in either whole tops (R2 = 0.65) or diagnostic samples (tips of runners back to the fifth to sixth fully expanded leaf; R2 = 0.65). Critical value for diagnostic samples was 0. 16% S while for whole tops of Siratro the value varied with mathematical model from 0.13 to 0. 15% S. Phosphate-extractable sulfate was the most effective soil test but irrespective of sampling depth, it accounted for less than 50% of the variation in Siratro relative yields. Whether sampled at 0- 10 cm or 0-90 cm, a critical range of 3-5 ppm phosphate-extractable sulfate was indicated for the Siratro component. Grass and total pasture relative yields were not correlated with the chemical tests employed.
