Residual effects of phosphorus fertiliser in a stylo-native grass pasture on a duplex red earth soil in the semi-arid tropics of North Queensland

Abstract

A field experiment was established to define the phosphorus (P) requirement for establishment and maintenance of a mixed legume pasture (Stylosanthes scabra cv. Seca, S. hamata cv. Verano, S. guianensis cv. Graham, Macroptilium atropurpureum cv. Siratro) introduced into a native grass pasture on an infertile duplex red earth. Rates of 0, 5, 10, 20, and 40 kg P/ha were applied to separate plots in year 1 (1982), 2, and 3. In year 5 (1986 growing season), half of each plot that had received 20 and 40 kg P/ha in year 3 was refertilised at the original rate to ensure that maximum yields were defined. Bicarbonate- or acid-extractable soil P concentrations of 8 mg/kg were sufficient for 80% maximum legume yield. The residual value of applied P in the surface soil, as measured by soil analysis, decreased exponentially, but an initial application of 40 kg P/ha was still sufficient to produce near-maximum legume yield after 5 years. Phosphorus application increased the dry matter yield of legume. During the establishment phase (years 1 and 2 after planting) yields reached maximum at 10 and 20 kg P/ha, respectively, but increased linearly in subsequent years. When the original rates were reapplied in year 5, peak yield occurred at 20 + 20 kg P/ha, and there was no difference between this yield and that from plots receiving 40 kg P/ha in year 1. Native grass yields increased with P application only in years 4 and 5 of the experiment. Stylos demonstrated good tolerance to low P supply. In year 1, 80% of the total legume yield consisted of Graham stylo and Siratro, whereas in subsequent years, Seca and Verano made up 70 and 20%, respectively, of the total, irrespective of treatment. Yield of legume at nil P, relative to maximum, increased from 5% in year 1 to 42% in year 5.