Phosphate rocks for direct application - which, where and why. Comment on 'The effectiveness of rock phosphate fertilisers in Australian agriculture: a review' by M.D.A. Bolland, R.J. Gilkes and M.F.D'Antuono (Aust. J. Exp. Agric. 1988, 28 655-668)

摘要

Although the review by Bolland et al. (1986) concludes thatnthe direct application phosphate rock (DAPR) has littlenpotential in Australian agriculture, a closer examination ofnAustralian research, both that included in and omitted fromntheir review, leads to very different conclusions. The reviewnmay be criticised on the following grounds:n1. A considerable amount of important work conductednoutside of Western Australia is omitted from the review; fornexample, the long-term grazed pasture trials at the WaitenInstitute and the Kybybolite Research Station (e.g. Cook 1939)nthat began in the 1920's. While unreplicated, these trials arennevertheless of real scientific value because of their long-termnnature. The massive differences that developed betweennfertilised and unfertilised plots over time tend to makenstatistical analysis largely superfluous. In addition, thencriticism by Bolland and Gilkes that insufficient fertiliser ratesnwere used in many trials is invalid, provided there is at leastn1 rate of superphosphate giving higher yields, as is the case innthese trials. At the Waite Institute, a trial on natural pasturenreceiving 500 mm rainfall showed that unreactive phosphatenrock (PR) gave higher production than that obtained withnsuperphosphate. On improved pasture (also 500 mm rainfall),nbut at a higher soil pH of 6.1 (water), a reactive PR equallednthe performance of superphosphate over a 10-year period;nunreactive PR (Christmas C) was about 60% as effectivenagronomically, due in part to the sulfur deficient status of thensite (basal S was not applied).nThe important work with Christmas A by K. D. McLachlannin the Australian Capital Territory in the 1950s, while listed innthe References, is not mentioned in the review. This isnsurprising, because McLachlan (1960) describes results fromn2 multi-rate 6-year pasture trials (6 replicates) at soil pHn(water) of 5.6 and 5.8 and higher average rainfalls of 750 mm.nHe concluded that: 'At equivalent levels of applied phosphorus,nsuperphosphate was a better fertiliser than rock phosphate innthe early years of the trial. Rock phosphate and superphosphatenboth had a good residual value, all the yields obtained withnboth fetilisers were not significantly different after the first fewnyears. The total yield of pasture over the 7-year period wasnsimilar for both fertilisers.'