PRINCIPAL COMPONENT ANALYSIS (PCA) APPLIED TO EXPERIMENTAL DATA OF A BIOLOGICAL PROCESS TO CLEAN UP DIESEL OIL-POLLUTED SOILS

In several scientific studies, the Principal Component Analysis (PCA) is used to process experimental data. The PCA aims to reduce the dimensionality of a data set consisting of many interrelated variables, preserving the variation as much as possible. This is achieved by transforming that set to a new one, that contains the principal components (PCs), which are uncorrelated, and which maximize the variance of the original data. In the present study, the PCA was applied to experimental tests on aerobic bioremediation of soil polluted with diesel oil. Twelve microcosms (200 g of soil) were set up, polluted with commercial diesel oil (70 g/kg of dry soil), and hydrated with a mineral salt solution to stimulate the indigenous bacterial consortia. The aerobic degradation was studied changing two operative parameters: water content (u%) and carbon to nitrogen ratio (C/N). Three different values of water content (u% = 8%, 12% and 15% by weight) and four of C/N (60, 120, 180 and 300) were tested. The microcosms were monitored for 30 days and these parameters were measured: 1) the CO2 production, to evaluate the microbial respiration, 2) the fluorescein production, to know the microbial activity, and 3) the diesel oil concentration, to check the pollutant degradation. The PCA analysis was done considering the water content values as variables and the C/N ones as systems to be analyzed. By a linear transformation, two principal components (PC1 and PC2) were determined. The experimental data of CO2 production, fluorescein production, and diesel oil concentration were used, and overall variances always over 94% were obtained. The results showed that two microcosms stood out from the others, namely that with u% = 12% b.w. and C/N = 180 since it produced the highest amount of CO2 and fluorescein, and that with u% = 8% b.w. and C/N = 120 due to its highest diesel oil removal efficiency.