EFFECTIVENESS OF DIFFERENT METHODS OF APPLYING SUPERPHOSPHATE FOR LUPINS GROWN ON SANDPLAIN SOILS

In 3 experiments in 1991 on very sandy soils near Eradu, Western Austr alia, the effectiveness of superphosphate for producing lupin (Lupinus angustifolius L.) seed (grain), was measured for fertiliser applied a t 0-73 kg P/ha to the soil surface just before sowing (topdressed), or banded with the seed, or 8 cm below the seed while sowing 5 cm deep. At all sites, banding phosphorus (P) below or with the seed was equall y effective as applying P to the soil surface. In a fourth experiment, on a very sandy soil near Badgingarra, Western Australia, levels of P (0-547 kg P/ha) as superphosphate, had been applied once only from 3 to 7 years previously (1985-89). The P applied in previous years was f ound to have leached. In 1992, superphosphate (0, 9, 18 and 36 kg P/ha ) was applied across all the original plots. Fertiliser was either app lied to the soil surface just before sowing lupins, or banded with the seed at 5 cm depth or at 8 cm below the seed. Grain yields from bandi ng P below the seed exceeded those where P was topdressed when <250 kg P/ha had been applied in previous years, or where the Colwell soil-te st P for the 10-20 cm depth was <10-15 mg P/g soil,When >250 kg P/ha h ad been applied in previous years, sufficient P had leached well below the seed, so there was little response to P and no advantage in placi ng freshly applied P below the seed when sowing. A possible explanatio n for the different results at Eradu and Badgingarra is provided.