A comparison of two phosphorus soil tests as inputs to a pasture growth model : a thesis presented in partial fulfilment of the requirements for the degree of Master of Agricultural Science in Soil Science at Massey University

Abstract

Glasshouse and field studies were carried out to investigate relationships between plant growth and extractable soil phosphorus and between fertilizer phosphorus and extractable soil phosphorus respectively. The purpose of the studies was to provide information with which to quantify the parameters of a simple model designed to predict relative pasture yield as a function of soil and fertilizer phosphorus. The relationship between yield and water-extractable soil P differed markedly between two soils of different P retention properties in glasshouse studies using both intact cores and conventional pots. To illustrate this difference, the levels of water-extractable P (0-8 cm depth) in intact cores required for 90% of maximum yield were 12.7 and 2.6 μg/g soil in the soils of lower and higher P retention respectively. In contrast, the relationship between yield and Olsen (bicarbonate-extractable) P was much less soil type dependent. The corresponding levels of Olsen P in intact soil cores required for 90% of maximum yield were 17.7 and 17.8 ug/g soil respectively. For modelling purposes, the Olsen procedure was therefore considered to provide a more suitable index of plant available soil P from which to predict pasture production on soils differing in P retention. The proportion of yield variation accounted for by differences in extractable soil P was 25% or less in initial harvests from the intact cores, 50-75% in later harvests from the intact cores and 89-97% in the pot experiments. The results of the intact core experiments, however, were considered to be more directly applicable to the field situation than were the results of the pot experiments. Seasonal changes in extractable soil P in Tokomaru silt loam included an increase during the dry season to reach a peak in late autumn followed by a decline in winter. The magnitude of these changes with respect to Olsen P was approximately 2.5 and 5 μg/g soil in the 0-8 cm and 0-4 cm depths respectively. A subsequent decline in extractable soil P during the spring and second summer was attributed largely to plant uptake of soil P and its loss in discarded clippings. The application of superphosphate increased extractable soil P in proportion to the rate applied. The increases per unit of applied fertilizer P, in both absolute terms and relative to an initial (time-zero) increase, were greater in a soil of low P retention (Tokomaru) than in a soil of high P retention (Ramiha). Water-extractable P (0-8 cm depth) was increased on average by 2.3 and 0.2 μg/g in the Tokomaru and Ramiha soils respectively six months after the application of 40 kg P/ha as super-phosphate. The corresponding average increases in Olsen P (2.7 and 1.1 μg/g) were greater, and differed less between the soils, than the increases in water-extractable P. Thus, neither soil P extraction procedure was independent of soil type in terms of the effects of applied fertilizer P. For modelling purposes the effects of applied fertilizer would need to be assessed in a wider range of soils. The level of water-extractable P in stored, air-dry soils was found to undergo short-term fluctuations, apparently due to changes in the conditions of extraction such as variations in the pH of distilled water. Longer-term increases of 25-100% in the level of water-extractable P of stored soils also occurred. No reason for the latter changes was apparent.