Herbage and animal production from native pastures and pastures oversown with Stylosanthes hamata. 2. Modelling studies

Abstract

Data from a long-term grazing experiment were used to develop relationships suitable for modelling between-year variation in pasture and animal production. The experiment was conducted near Townsville in North Queensland and consisted of a factorial combination of 2 stocking rates (0.6, 1.2 steers/ha) x 2 pasture types (native pasture, native pasture plus Stylosanthes hamata cv. Verano) x 2 superphosphate rates (nil, 300 kg/ha.year). Cattle were weighed monthly, and the herbage presentation yield and stylo content estimated annually. Relationships between annual liveweight gain (LWG) and a range of climate-derived parameters including the number of 'green days' were calculated from a water balance and pasture growth model. When a different intercept was allowed for each pasture and stocking rate combination, there was a strong relationship between annual LWG and the number of 'green days' (R2 = 0.78). In a more general relationship, LWG was related to green days, stylo content, and utilisation rate (R2 = 0.58). No relationships was found that would enable reliable prediction of between-year variation in the pasture attributes of stylo content and the water use efficiency of herbage growth. The times of the year when daily rates of LWG changed could be predicted satisfactorily, but rate of gain within each phase varied considerably from year to year. Differences in LWG between stocking rates occurred when there would have been restricted amounts of new green feed, or of stylo in the late wet season.