PHOSPHATE FERTILIZER RESIDUES IN WHEAT-GROWING SOILS OF THE WESTERN-DOWNS, QUEENSLAND

Effects of repeated annual application and residual effects of past ph osphorus (P) applications were studied at Billa Billa and The Gums, Qu eensland, on 2 Vertisols representative of a large cropping area in th is region. Phosphorus was applied annually at 0, 4, 8, 12, and 25 kg/h a at sowing to each wheat crop between 1978 and 1988. Phosphorus was a lso applied at 25, 50, 100, 200, or 400 kg/ha as a once-only applicati on, before wheat was sown in 1978, and was incorporated by tillage. Po t experiments were conducted each year with soil (6 kg/pot) gathered f rom field plots in May before wheat sowing to determine the residual v alue (RV) of previously applied P fertiliser under standard conditions of frequent watering. In 1979 the RV was determined in a second serie s of pots with a water deficit to simulate conditions of plant moistur e stress that were to dominate the field experiments. The RV of P fert iliser was reduced to one-third that determined under well-watered con ditions. Low residual values (8-47%) determined in the field in 1979 f or P rates 25 and 50 kg/ha applied in 1978 were probably due to reduce d P uptake from dry topsoil throughout which P residues were incorpora ted by tillage. By contrast, P applied annually at a low rate (4 or 8 kg/ha) in a band close to seeds created profitable responses when eval uated over the sequence of successive wheat crops. The bicarbonate-ext ractable soil test value (Pbi) on either soil fertilised annually for n years with P at a rate R (kg/ha . year) was described by the model P -bf = P-bu + 0.29R Sigma 5{5 + (n - 1)}(-1) where P-bu is the soil tes t value (mg/kg) of unfertilised soil.