Interactions among Ca2+, Na+ and K+ in salinity toxicity: quantitative resolution of multiple toxic and ameliorative effects

Root elongation by wheat seedlings (Triticum aestivum L. cv. Scout 66) was not inhibited by NaCl or KCl up to 130 mM in culture solutions or by high Na+ (2 mg g−1 FW) or K+ (4 mg g−1 FW) in the root tissue, provided that [Ca2+]>2 mM in the rooting medium. At [NaCl], [KCl], or [mannitol] >250 mOs, root elongation was progressively inhibited, irrespective of high [Ca2+]. In contrast, shoot elongation was sensitive to any diminution of water potential, and Ca2+ alleviated the toxicity only weakly. At solute concentrations <250 mOs, the following interactions were observed. Ca2+ alleviated Na+ and K+ toxicity to roots by at least three separate mechanisms. K+ was more toxic to roots than Na+, but Na+ was more toxic to shoots. Low levels of K+ relieved Na+ toxicity, but low levels of Na+ enhanced K+ toxicity. Tissue concentrations of Na+ were reduced by Ca2+ and K+ in the rooting medium, and tissue concentrations of K+ were enhanced by Ca2+ and Na+. Several hypotheses relating to salinity toxicity can be evaluated, at least for wheat seedlings. The osmoticant hypothesis (salinity intoxication occurs because of diminished water potential) is true for shoots at all salinity levels, but is true for roots only at high salinity. The Ca2+-displacement hypothesis (Na+ is toxic because it displaces Ca2+ from the cell surface) is correct, but often of minor importance. The K+-depletion hypothesis (Na+ is toxic because it causes a loss of K+ from plant tissues) is false. The Cl−-toxicity hypothesis (the apparent toxicity of Na+ is induced by associated Cl−) is false. The results indicate that, apart from osmotic effects, high levels of Na+ in the rooting medium and in the tissues are not toxic unless Ca2+ is also deficient, a condition probably leading to inadequate compartmentation and excessive cytoplasmic accumulation. This study related growth to ion activities at plasma-membrane surfaces. These activities were computed by a Gouy-Chapman-Stern model then incorporated into non-linear growth models for growth versus toxicants and ameliorants.