Principal Component Analysis (PCA) to the study of levels and congeners distribution of PCBs in soils aPrincipal Component Analysis (PCA) to the study of levels and congeners distribution of PCBs in nd sediments from the southwest of Buenos Aires province

ALVAREZ, MÓNICA; KUKUCKA P.; TOMBESI NORMA; METZDORFF, AMÉRICA; AUDY O.; POZO KARLA; PRIBYLOBA, PETRA; KLANOVA J.

Puerto Madryn, Chubut, Argentina

Congreso; IV Reunión Argenyina de Geoquímica de la Superficie; 2016

CESIMAR-CENPAT-CONICET

PCA was used in order to deepen the results obtained for 7 PCBs congeners analyzed in four surface sediments collected from the northern shore of the Bahía Blanca estuary (S1 to S4) and in nine soils from different locations of the Bahía Blanca city (B1 and B2) and the nearby region (R1 to R7), about 100 km around the city, on the Southwest of Buenos Aires Province. The PCB congeners studied (PCB-28, -52, -101, -118, -138, -153, and -180) are considered as indicators due to their relatively high concentrations in technical mixtures and their wide chlorination range (3-7 chlorine atoms per molecule) (Webster et al., 2013). Results showed that the two first principal components explained the 91.07% of the total variability present in the data set. PC1, which explains 69.16% of the variability, characterizes the separation between soil samples of Bahía Blanca city from both urban and industrial zone (B1 and B2) + sediments from the Bahía Blanca estuary, next to port and industrial activities (S2 to S4) and soils from different sites located in the southwest of the Buenos Aires Province (R1 to R7), while the sediment more away from the industrial/urban area (S1) remain in the axis. On the other hand, the PC2 axis, which explains 21.91% of the total variability, draw a distinction between soils from urban and industrial zone of Bahía Blanca city + sediment next to an industrial discharge point (S3) (scored positively) and the rest of sediment samples of the estuary (scored negatively), while the rest of the samples do not exhibit a marked differentiation on PC2. The biplot graphic representation of the PC analysis also reveals the relationship between congener composition of PCBs (represented by number of Cl) and sample locations. All the PCBs congeners were clustered positively on PC1, and the variability is observed on PC2: the content of 7-Cl and 3-Cl PCBs congeners were clustered negatively and linked to soil samples S2 and S4 and, and 6-Cl and 5-Cl PCBs congeners clustered positively and linked to B1, B2 and S3 samples, while 4-Cl congener do not present a significant contribution to differentiate the studied samples. PCA was used as a first tool in order to evaluate possible associations between the origin of PCBs and samples sites. Taking into account the wide variety of sources of PCBs congeners, i.e. direct inputs, land runoff or atmospheric deposition, and different degradation pathways, probably studies with bigger data sets will be necessary to reach clearer conclusions.