Evaluation by Multivariate Statistical Methods of Relationship with Physicochemical Properties of Humic Acid and Some Soil

Humic acid is a very important component for soil. In particular, It is an important factor for regulation of plant nutrients. In this study, HA (humic acid) contents of soil samples taken from 12 different locations were determined by appropriate method. After extracting the humic acid from the soil samples, the remaining soil elements concentrations were determined by ICP-OES method. In addition, the element concentrations in the humic acid fraction were determined by the same method. In soil samples were determined organic matter (OM), pH, EC, clay, silt and sand. The relationships between all variables were evaluated by multivariate statistical analysis such as correlation, simple linear regression, PCA (Principal Component Analysis). Spearman rho coefficient was taken into consideration in the correlation analysis. Simple linear regression analysis showed significant regression model between HA and OM, pH, HAfMn (HAf: humic acid fraction), HAfSe, HAfP, TfCd (Tf:soil fraction), TfMn, TfPb, TfSe. In the PCA analysis, 4 factors were found to explain 88.23% of the total change. P1:OM, TPb, TSe, HA  P2:TAs, TSb, TBa, TMn, TFe  P3: TCd, TP, pH, TSn, Silt, TAl  P4: Kil, sand, TBe, EC. In the correlation analysis (N: 12), HA was significant positive with TfSe (r= 0.615, P <0.05), and significant negative with HAfBe (r= -0.786, P <0.01) and significant positive with HAfMn (r= 0.918, P <0.01) and significant negative with HAfSn (r= -0.700, P <0.05) and significant positive with HAfSe (r= 0.704, P <0.05) and significant positive with HAfP (r= 0.700, P <0.05) correlation was found.