The availability and distribution of an element are determined by soil conditions and the geochemical fractions. In this study, canonical correlation (CCA) was used to examine the associations between the soil parameters (X set) of the Amik plain and the fractions of selenium (Se) (Y set). A statistically significant correlation (P < 0.01) of 0.391 ± 0.06 was observed between the first canonical variable pair (FCVP). The clay, sand, and silt variables exhibited the greatest standardized canonical coefficients, with values of 74.55, 75.83, and 94.77, respectively. Additionally, the B3 percent had a standardized canonical coefficient of 1.029. The analysis of the FCVP showed that the original variables of the X set and/or its canonical variable (U1) had the strongest linear connection (0.770) with organic matter (OM) and the B3 percentage in the Y set (0.603). The OM (0.3012) exhibited the most substantial linear structural correlations between the canonical variables of the Y set and the original variables of the X set. However, the canonical redundancy analysis revealed that in the FCVP, the variance of the X set was described by its canonical variables to the extent of 20.97%, whereas the Y set was explained by 21.96%. Hence, it can be inferred that the use of canonical correlation is effective in elucidating the relationship between soil parameters and various components of plant nutrition.