EFFECT OF LIME AND FORM OF SUPERPHOSPHATE ON PRODUCTIVITY OF DRYLAND PASTURES

The effect of initial high lime (7500 kg ha(-1)), annual low lime (200 kg ha(-1) yr(-1)), and no lime (all with 200 kg ha(-1) yr(-1) superph osphate) and dicalcic superphosphate (400 kg ha(-1) yr(-1)) was compar ed over six years under sheep grazing and mowing on a low rainfall (80 0 mm), low pH (5.4), phosphorus (P) responsive (Olsen P 5) Matapiro ye llow-grey earth (Duric palic) soil near Hastings, Hawke's Bay, New Zea land. High lime significantly increased pasture dry matter yield compa red with other treatments in the first three years of the trial under both grazing and mowing. There were no differences in annual yield bet ween the other treatments. The response to high lime in relation to no lime in the first 3 years was greater under grazing (21%) than mowing (12%). There was a greater number of legume plants per m(2) under mow ing than grazing. The increase in pasture yield from high lime in the grazing trial was because of an increase in grass rather than legume y ield. Legume yield averaged over all years was similar for the other t reatments. There was no significant difference in the P concentration of grasses between treatments. Compared with low lime, high lime resul ted in an initial increase of 0.9 pH units. Low lime and dicalcic supe rphosphate maintained the initial soil pH. The initial low Olsen P lev els were only slightly increased by the low rates of P fertiliser used . After six years, P applications ceased on the dicalcic and low lime treatments, and the high lime plots were split into no P and P (20 kg ha(-1) yr(-1)) applied. During the residual effect phase (Year 7-17), there was no difference in annual yield between no lime and high lime when P applications continued. For all treatments, where P was withhel d, annual yield was lower than where P was maintained. Application of high initial rates of lime in this environment resulted in a short-liv ed (3 year) response in pasture yield despite a large increase in soil pH. At equivalent rates of P there was no difference in the effective ness of single and dicalcic superphosphate. The liming effect of dical cic was matched by applying the equivalent low rate of lime as a dry m ix in addition to superphosphate.