ABSTRACT
Summary Fine-stem stylo (FSS), Stylosanthes guianensis var. Intermedia, is a perennial legume suitable for sowing into native Heteropogon contortus pastures on well-drained soils in southeastern Queensland. A grazing system is described that aims to maintain a productive FSS pasture and grow steers to 480 kg live weight at 27 months of age. This goal requires gains of 300 kg/steer in 20 months after weaning. Results from a short-term field study of the system were analyzed for long-term repeatability using a simple model that predicted animal production from stocking rate and weather records. The system was studied near Gayndah, Queensland, on 11.3 ha of FSS and H. contortus pasture invaded by Rhynchelytrum repens. Each June after weaning, seven steers (five-eights Hereford, three-eights Sahiwal) commenced a 20-month grazing cycle. Thus, two animal classes grazed together from June to January. From February to May, after the older animals were slaughtered, the stocking rate was halved to allow FSS to grow before winter. The mean annual stocking rate was 1.0 steer/ha. Since 1976, three grazing cycles have been completed with total live-weight changes of 276, 326, and 315 kg/steer to give carcasses suited to local trade. FSS contributed 10% of pasture yield, but weeds progressively increased. The animal production model predicted live-weight change (LWC) in winter, spring and summer, and autumn from pasture-growth index/head, the mean-growth index (GIX) for a season divided by stocking rate (SR). GIX was the product of separate indices (range 0 to 1) for temperature, solar radiation, and soil water supply. The regressions predicting LWC from GIX/SR were derived from results of the field study and two published studies; they accounted for 85% to 89% of variation in LWC. Complete weather records (1958-1980) were processed to estimate GIX from seasonal rainfall and GIX in the previous season. This enabled LWC to be simulated from monthly rainfall from 1870 to 1980. Three mean annual stocking rates (0.74, 1.0, and 1.3 steers/ha) were considered over 109 grazing cycles. Frequency distributions of total live-weight gain showed that SR affected repeatability. The model estimated long-term repeatability from a restricted data base but ignored possible influences of extreme feed shortages and changes in soil fertility and botanical composition of LWC. It supported the field study by indicating that a reduction in SR would improve repeatability of production and reduce weeds; it provided information that would otherwise have to come from a long-term grazing trial; it suggested that pasture stability in drought should be studied; and it indicated that the grazing system is suitable for commercial application.
