New statistical approach for the interpretation of nutrient interrelationships. V. Zinc/iron

Abstract

Maize (Zea mays L. c.v. Pioner) plants were grown in a controlled greenhouse in plastic pots (25 plants/pot) containing 25g of soil and 475g of quartz sand. Twelve different soils were tested, each one repeated five times. Dry weight and nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), iron (Fe), copper (Cu), manganese (Mn), zinc (Zn), and boron (B) concentrations were determined separately for shoots and roots. Based on multiple linear correlation analysis and determination coefficients (R), a statistical method was developed to select the most effective variable to explain the variation of the dependent veriable. The contribution of every selected variable was then calculated and expressed as percent. In roots Zn was affected by N (55%) and negatively by B (‐22%), while Fe was dependent on N (7%), Ca (30%) and Mn (33%). In the shoot, Zn was dependent only on B (53%), while Fe was dependent on N (10%), Ca (40%), Cu (37%) and Mn (‐1%). For the whole plant, these relations changed, as Zn was dependent only on P (39%), while Fe was dependent on N (13%), Ca (20%) and Mn (35%). Thus, we conclude that Zn facilitates the Fe/Mn relation, as well as those of other nutrients.