Analyses of statistical transformations of row data describing free proline concentration in sugar beet exposed to drought

Abstract

Eleven sugar beet genotypes were tested for their capacity to tolerate drought. Plants were grown in semi-controlled conditions, in the greenhouse, and watered daily. After 90 days, water deficit was imposed by the cessation of watering, while the control plants continued to be watered up to 80% of FWC. Five days later concentration of free proline in leaves was determined. Analysis was done in three replications. Statistical analysis was performed using STATISTICA 9.0, Minitab 15, and R2.11.1. Differences between genotypes were statistically processed by Duncan test. Because of nonormality of the data distribution and heterogeneity of variances in different groups, two types of transformations of row data were applied. For this type of data more appropriate in eliminating nonormality was Johnson transformation, as opposed to Box-Cox. Based on the both transformations it may be concluded that in all genotypes except for 10, concentration of free proline differs significantly between treatment (drought) and the control.

Keywords

• tolerance
• water deficiency
• sugar beet
• proline
• statistical transformation
• Box-Cox transformation
• Johnson transformation