Application of the Plackett-Burman Model on Soil Fertility Determinants

Abstract

In the present study, a soil from the region of Fez-Sefrou Morocco was screened for some physicochemical characteristics using the Plackett-Burman model in order to know the most important factors that promote its fertility. Five independent variables were selected: pH, electrical conductivity, humidity, organic matter, C/N ratio. These variables were evaluated by statistical analysis, based on their significance, the value of the coefficient of determination and the Pareto chart. The results suggest that humidity and C/N ratio, have an influence with a high level of confidence. While the other three show no significant effect on the content of nutrients in the soil. Analysis of the R2 variance value also showed that the models used for prediction were large and significant factors (p less than 0.05). Pareto chart plots for each response and its characteristics provided accurate data to select well-fitting variables for further optimization.

Keywords

• plackett-burman design
• pareto chart
• variables
• screening
• coefficient of determination