Superphosphate and stocking rate effects on a native pasture oversown with Stylosanthes humilis in central coastal Queensland. 3. The balance of native species in the pasture

Abstract

The percentage frequency of all herbaceous species in a native pasture oversown with S. humilis (T.S.) was measured at the beginning and end of an eight-year grazing experiment. Treatments imposed were three levels of molybdenized superphosphate (nil; 125 kg ha-1 annually; 250 kg ha-1 annually plus an extra 250 kg ha-1 initially) in factorial combination with four stocking rates and two sowing methods for T.S. (ground sowing into spaced cultivated strips or aerial sowing). Stocking rates were gradually increased during the experiment and for the last three years overlapping ranges were used for the three fertilizer levels; the overall range was then from 0.55 to 1.65 beast ha-1. Treatments did not induce any major changes in the proportions of native species in the pasture. Heteropogon contortus remained dominant and its frequency increased in all treatments. Of the total of 113 species recorded, only 18 reached 10 per cent frequency in at least one paddock on one occasion. Several of these showed statistically significant changes in response to treatment, but their relative importance was not greatly changed. Examination of the results by principal co-ordinate and canonical analysis, showed that the first vectors of the two data sets were highly correlated, indicating that the major control on floristic composition was not changed by oversowing with T.S. or by the treatments imposed. Treatment effects dominated in the second and third vectors in the final year. The controlling vector was found to be highly correlated with soil calcium, potassium, total exchange capacity and organic carbon, but the basis for the correlations remains undetermined.