Field and laboratory studies of the movement and reactions of phosphorus in soils : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Soil Science at Massey University

Abstract

Low and stable concentrations of phosphorus (P) forms and sediment were obtained in stream flow from two small, adjacent, scrub-covered, and minimally-disturbed catchments near Palmerston North, New Zealand. In contrast, higher and irregular concentrations and loadings were obtained following land clearing, P fertilizer application, and the establishment of grazed pasture. The need for intensive stream sampling, as well as complete hydrograph data in order to obtain reliable information on stream loadings, is emphasised. High and fluctuating concentrations of P forms and sediment were obtained following the change in land use. A high proportion of the P and sediment loss occurred in the storm runoff component of stream flow. The estimated losses of fertilizer P in stream flow (approximately 1% of that added) were very small from an agronomic standpoint but they represent large proportional increases in the loadings of P forms in stream flow. The high amounts of water-extractable P present in the soils of the catchment (field soils), immediately following the aerial application of fertiliser P, declined rapidly to lower, more stable values. This pattern of decline for field soils was replicated using small pots established in the field (pot soils) and containing fertilized soil representative of the catchments. Close correlations were obtained between water-extractable P in the upper 1cm of field and pot soils, and mean dissolved inorganic P (DIP) concentrations in the surface runoff component of stream flow in closely-following storms. The possibility of predicting DIP losses in surface runoff from soils using a water-extraction technique is thus indicated. The decrease with time in the amounts of water-extractable P observed after superphosphate addition to field and pot soils was reproduced in the laboratory. This relationship validated the use of laboratory studies to examine the rate and extent of interaction of fertilizer P occurring in field soils and to predict the potential movement of fertilizer P from soils to waters. The decline in water-extractable P closely paralleled the decrease in plant uptake of P with time following fertilizer P addition to two constrasting soils. This suggested that water extraction may be a useful soil-testing procedure for predicting P availability to plants, as well as the movement of P in surface runoff from soils. The rate of decline in water-extractable P in a given soil was proportional to both the amount of P added and the amount initially extractable immediately following P addition. This suggests that the rate and extent of P sorption in a soil is directly related to soil solution P concentration. Differences were obtained, however, between three contrasting soils in the relative rate and extent of P sorption. A kinetic model based on the Langmuir equation was developed to simulate the decline in water-extractability of P added to three soils. Three populations of sites were assumed and the appropriate sorption maxima and binding energy constants were derived from sorption isotherm studies. The model provided a satisfactory prediction of the fate of different amounts of fertilizer P. It is probable that the further development of this model would provide a useful basis for predicting the fate of P added to soils and the potential movement of added P in surface-runoff waters.