Evaluation of management options for increasing the productivity of tropical savanna pastures. 3. Trees

Abstract

The effect of killing trees upon the production and quality of native perennial grasses, Themeda triandra, Chrysopogon fallax, Sehima nervosum, and Sorghum plumosum, and oversown legumes from the genus Stylosanthes, was studied over 4 years at Katherine, in the semi-arid tropics of northwestern Australia. The pastures were either unfertilised or received low inputs of superphosphate, and for each fertility level were grazed at 3 stocking rates. At no time were legume yields affected by killing the trees but, in the first 3 years, the amount of grass was approximately twice as much when the trees were killed. During this period the mean grass yields declined 4-5 fold from about 2.2 t/ha. By the fourth year the advantage from tree killing upon grass yield was apparent only at the lowest stocking rates at each fertility level. Nitrogen concentrations of the grasses and legumes, with the exception of S. hamata, were increased 7 and 10% respectively above the mean annual values of 0.89 and 1.75% where the trees were killed, while the phosphorus and sulfur concentrations were not affected. Tree killing had no effect upon wet season liveweight gains during the last 2 years of the experiment. However, there were some benefits during the dry season when weight losses were lower for most treatments during the early dry season (June-September) and also lower for the lowest stocking rate treatment without fertiliser during the late dry season (October-November).