Spatial dependency of arsenic, antimony, boron and other trace elements in the shallow groundwater systems of the Lower Katari Basin, Bolivian Altiplano.

Spatial patterns, cluster or dispersion trends are statistically different from random patterns of trace elements (TEs), which are essential to recognize, e.g., how they are distributed and change their behavior in different environmental processes and/or in the polluted/contaminated areas caused by urban and industrial pollutant located in upstream basins and/or by different natural geological conditions. The present study focused on a statistical approach to obtain the spatial variability of TEs (As, B and Sb) in shallow groundwater (GW) in a high-altitude arid region (Lower Katari Basin, Bolivian Altiplano), using multivariate analysis (PCA and HCA), geochemical modeling (PHREEQC, MINTEQ) and spatial analyses (Moran's I and LISA), considering the community supply wells. The results indicate that despite of the outliers there is a good autocorrelation in all cases, since Moran's I values are positive. The global spatial dependence analysis indicated a positive and statistically significant spatial autocorrelation (SA) for all cases and TEs are not randomly distributed at 99% confidence level. The results of hydrochemical modeling suggested the precipitation and stability of Fe (III) phases such as goethite. The re-adsorption of As and Sb on the mineral surface in the aquifer could be limiting the concentrations of both metalloids in southern regions. Spatial autocorrelation was positive (High-High) in northwestern (arsenic), southeastern (boron) and northeastern (antimony) region. The results reflected that the As and Sb are the main pollutants linked to the natural geological conditions, but B is a main pollutant due to the anthropogenic activities. Furthermore, >50% shallow groundwater exceeded the WHO limit and NB-512 guideline values for Sb (87%), B (56%) and As (50%); therefore the spatial distribution and concentrations of these TEs in GW raise a significant concern about drinking water quality in the study area.