Use of Multivariate Statistics Methods to Determine Grain Size, Heavy Metal Distribution and Origins of Heavy Metals in Mersin Bay (Eastern Mediterranean) Coastal Sediments

Abstract

The aim of the study was to determine variability, heavy metal content and potential origins of heavy metals of the sediments in a total of 60 locations representing coastal sediments of Mersin Bay. Grain size distribution and heavy metal contents were measured and multivariate statistical analyses were performed on obtained values. In grain size distribution, Oz-4, 5, 6, 11, 12, 13, 14, 15, 16, 18, 19, 22, 23, 46, 48, 49, 50 stations showed a bimodal distribution. This distribution developed due to river networks near the locations. Heavy metals are sequenced as Cr, Mn, Sr, Ni, V, Zn, Co, Zr, Rb, Ce, Cu, Sc, Li, Y, Pb, As, Nb, Mg, Fe, Al, Th and U from the higher value to the lower value according to their abundance. According to frequency histogram, Ni, Fe, Al, which showed the highest concentration values among heavy metals came from short-medium distance; while Cr, Ti and Mn came from short distance. Based on these findings, it should be thought that heavy metal sources affecting study area are in short distances to study area. All the elements were represented with three sector principal component analysis. Total variances of Pb, As, U, Th, Sb, P, La, Ba, Na, K, W, Ce, Li and Rb elements which represent (F1) factor were explained by 30.591 %. Variance of Mn, Fe, V, Ti, Al, Zr, Y, Sc and Hf elements which represent (F2) factor were explained by 18.749 %. Total variance of Mo, U, Sr, Ca, P, Ti, Nb and Ta elements which represent the third factor (F3) were explained by 14.512 %. These data are significantly consistent with the dendongram prepared according to coefitic correlation coefficients. Hierarchical group analysis dendongram showed that Q-type cluster had a 50 % arbitrary similarity level and that contamination generally occurred in group 3. It can be thought that similar groups had the same properties during contamination. In regression data performed according to Fe, "Model summary" (according to R² = 99.8 value) was significantly adequate for statistical data and "Anova" was highly reliable with 36 explanatory variables. Heavy metals in the study area such as Cr, Mn, Ni, Zn, Co, Cu, Pb, Mg, Al, Cd, Sb and Ti might show toxic effects. Heavy metals such as As, Ag, Fe, Mo and Sn should also be paid attention. Al, Fe, Ti, Mn, Cr, Ni, Co, Pb, Zn and V showed an anomaly according to Kizkalesi and Susanoglu coastal sand. Cr, Ni, Co, Mg, Ti, Fe and Mn increased due to Mersin Ophiolite. The areas where basic/ ultrabasic rocks outcropped in the region can be considered as the source of natural contaminations. Anthropogenic factors, coastal sediments, coastal erosion and lithological effects are the main causes of contamination in the study area, which covers a very wide area. In addition, the port, river entrances, highway, urban wastes, tourist facilities and industrial sites increased the density of anthropogenic effect.