Tropical Grasslands (1992) Volume 26, 171–180

Effect of tree clearing and seedbeds on the establishment, growth and population dynamics of siratro, green panic and signal grass oversown into a speargrass pasture in south-east Queensland


1CSIRO Division of Tropical Crops and Pastures, Brisbane, Queensland, Australia
2CSIRO IPPP Biometrics Unit, Brisbane, Queensland, Australia


The high cost of establishing improved pastures over large areas using cultivated seedbeds, and the unreliability of surface sowing, have prompted research into alternative, low cost techniques of pasture establishment. Seed mixtures of siratro (Macroptilium atropurpureum) and green panic (Panicum maximum var. trichoglume) or siratro and signal grass (Brachiaria decumbens) were either broadcast into live speargrass (Heteropogon contortus) pastures, into dead speargrass pastures that had been sprayed with herbicide, or sod-seeded into live speargrass pastures with minimal disturbance. The 3 seedbeds were imposed on native pastures that had been cleared of trees for 6 years, had the trees killed at sowing or where live trees remained. The establishment, growth and survival of seedlings, together with yield and botanical composition of the swards, were studied over 4 years.
Pasture establishment was highest in the herbicide-treated seedbeds without live trees and lowest where seed was broadcast into undisturbed speargrass pastures where trees had been cleared 6 years before sowing, and in all seedbeds where live trees were present. In the herbicide seedbeds without live trees, lower levels of plant competition led to better seedling survival and resulted in pastures with a greater proportion of sown species. Even though siratro set seed each year, and there were a number of seedling recruitment events, after 4 years, over 97% of the siratro population was composed of survivors from the initial emergence.

Download full article (659 KB PDF)  

  Return to Contributed Articles