INTERPRETATION OF PHYSICO-CHEMICAL PARAMETERS IN SEDIMENTS AND WATERS OF THE NEGRO RIVER (ARGENTINA) BY MULTIVARIATE STATISTICAL ANALYSIS

PAOLA SUÁREZ; MARISOL VEGA; RAFAEL PARDO; ALICIA RONCO

Budapest

Conferencia; XIII Chemometrics in Analytical Chemestry Conference; 2012

Hungarian Chemical Society

Rivers are important sources of energy that actively modify the configuration of the landscape.Sediments and waters transport both nutrients and pollutants, so the knowledge of the parametersgoverning river dynamics is of vital interest. Qualitative and quantitative studies of the load and thedetrital materials transported in solution by rivers and of the dominant transport mechanism provideinformation for environmental assessment and for the determination of evolutionary trends in a region. The Negro river is located in Chaco province, Northeast Argentina, between 27 º 04’32 “S - 59 ° 27’23"Wand 27 º 25’42 ”S - 58 º 57’42" W. It is an autonomous course, typical of low slope areas, and fed by local rainfall, but in its final section is under the influence of the flood regime of the Paraná River, its final collector. Seasonal samplings of bottom sediments, suspended sediments and water were conducted semi-annually for three years at four river locations: headwater area (1), at middle stretch channelled (2, 3) and near mouth section (4). Each location was sampled for sediments at three points: centre of the course, and left and right margins. Laboratory tests consisted of physical methods (concentration of suspended solids and particle size) and chemical methods (identification of major anions and cations). ANOVA applied to some sediment parameters demonstrated that only variations caused by season and location were significant, whereas cross variation was not significant. Therefore, only central samples were considered for subsequent analysis. Exploratory data analysis was then performed by multivariate chemometric tools such as Principal Component Analysis (PCA) (two and three way) and Cluster Analysis. Preliminary results obtained by PCA and cluster analysis show a small number of latent factors with hydrogeochemical meaning. It is observed that the variables analyzed in water are significantly correlated with each other and with the contents of some ions readily mobilized from sediments. PCA and cluster analysis allowed also classify the samples according to these latent factors. It is observed that the variability between samples reflects both a component related to seasonal rainfall as the spatial component, with a clear differentiation between headwater sediments and sediments near the river mouth.