Behaviour of monoammonium phosphate in alkaline and calcareous sandy soils
Venter, Andrie Elize
MetadataShow full item record
Soluble phosphorus (P) in soil is subject to fixation in either low or high pH soils. A variety of other soil processes contributes to this process, for example clay mineralogy, organic matter, sesquioxides and carbonates. Arid and semi-arid areas which are high in carbonates require proper P fertilisation to ensure sustainable crop production. Proper P fertilisation is hampered by the choice of which agronomic soil P test is best employed for fertiliser recommendations. The main objectives of the study were, firstly, to establish the amount of monoammonium phosphate needed to increase extractable P in the upper Orange River catchment soils with different calcium and phosphorus contents. Moreover, relationships between the application of P and the P extracted by the Olsen, Bray 1, Mehlich 3 and the Ambic 1 methods were investigated in the study soils which are used for irrigation. Lastly, the study compared extractable P contents in these catchment soils, which were analysed by three analytical laboratories. Soil samples from the orthic A horizon were collected at six sampling sites in the upper Orange River water management area below the Vanderkloof dam in the southwestern parts of the Free State, and eastern parts of the Northern Cape. These samples were dried and sieved before conducting a two-month laboratory incubation study at room temperature, where they were treated with seven levels of monoammonium phosphate. During incubation, the samples were exposed to several wetting and drying cycles. The P in the soil samples was extracted with the Olsen, Bray 1, Mehlich 3 and Ambic 1 methods for colorimetric determination by the UFS laboratory. These samples were also analysed for extractable P by two commercial laboratories with the Olsen, Bray 1 and Mehlich 3 methods. Analyses of variance were conducted with IBM SPSS Statistics 25 at a 95% confidence level. The P content means of extraction methods and application levels were then compared with Tukey’s procedure, also at a 95% confidence level. Simple regression analyses were also done to meet the objectives of the study. The six sites can be catagorised as low calcareous (<0.7% calcium carbonate) with low to high P contents (8.9 to 24 mg kg-1 Olsen P) and as high calcareous (>3.3% calcium carbonate) with low to high P contents (2.4 to 42 mg kg-1 Olsen P). Thus phosphorus requirement factors (PRF) estimated from regression equations varied significantly between the four extraction methods. The Bray 1 method showed significantly unrealistic PRFs of 1.8 to 384.6 kg P ha-1. By contrast, the variation of the PRFs for the Mehlich 3 method was very slight (0.9 to 2.1 kg P ha-1). The PRFs of the Olsen method (4.6 to 6.1 kg P ha-1) and Ambic 1 method (1.7 to 4.3 kg P ha-1) were more in line with other studies. Relationships between applied P and extracted P showed that various regression equations fitted the data with different methods. Although almost linear, polynomial equations best described the relationship with R2- values exceeding 0.98 for the Olsen method. Poor relationships (R2- values less than 0.57) were regressed with Bray 1 method data at calcareous sites. A variety of equations fitted the data best when using the Ambic 1 method. Good relationships between the P extracted with these methods may have a positive influence on fertiliser recommendations when conversion of P contents is required. Some significant differences between extractable P for a particular method were observed by the three analytical laboratories. These differences can result in fertiliser recommendations bring offered with limited confidence. This study proved that the mineralogical, physical and chemical properties of a soil ultimately prescribed which method is the most suitable to extract P for reliable P recommendations. The Olsen method proved to be the most reliable on both the non-calcareous as well as the calcareous soils. The PRFs estimated with Olsen data ranged from 4.6 to 6.1 kg P ha-1. These values can serve as a basis for increasing extractable Olsen P to the required optimal levels. However, field studies are warranted to establish threshold values for fertiliser recommendations for each extraction method.