The influence of spatial soil variability on the effectiveness of Minjingu Phosphate Rock and crop performance was studied in four field trials in north- eastern Tanzania using the post -mortem residual analysis and the nearest neighbor means techniques. Residual analysis refers to the determination of deviation of a particular point from the regression line (predicted value). T he residual values were obtained by subtracting the predicted yield from the observed yield and t he predicted yield was calculated from the full statistical model, including all main effects and interactions . The field trials were located in two contrasting sites; one with a low pH (5.0) and the other with a high pH (5.8). Maize ( Zea mais (L.) var. TMV1) was planted during the long and short rains of 1997. We differentiated two different sizes of trials namely; large and small trials based on the number of plots and therefore area covered. Results showed that variation in maize yields was explained more by soil pH rather than soil available P. Soil pH influence d the relationship between P values determined by Bray -I and Olsen. Phosphorous values obtained by BrayI -P tended to be higher than those by Olsen -P in the low pH soil and lower than Olsen- P in the high pH soil. At low pH, BrayI -P was > 0.65 x Olsen- P. The refore, to be able to correctly identify the important factors influencing soil variability especially when P is involved, the choice of appropriate method of P determination in relation to the pH of the soil is important. Further, the two techniques for s tudying spatial variability (residual analysis and nearest neighbor means) were more effective in isolating environmental effects from treatment effects in the larger trials than the smaller ones . There was more variability in the first than the second season due to the uniform management and the treatments applied. Post -mortem residual analysis and the nearest neighbor means techniques should therefore be used in large trials to isolate environmental effects beyond what normal blocking can do.
Año de publicación
2011
Autores
Mowo G J; Janssen B H; Oenema O; Mrema J P
Idioma
English
Palabras clave
Crops, Soil fertility, Yields
Geográfico
Kenya, Tanzania, Uganda
