Table 1.
Deconstructing the experiment of Krishnasamy et al. [14] to determine the effects of K+ and Na+ into the actual ion levels.
The experiment varied three ions K+, Na+, and Cl-. The effects of K+ and Na+ are confounded with those of Cl-.
Table 2.
Ionic compositions of the “soil solution equivalent” (SSE) and ARS-Media for Excel formulation and how each compares to the ion levels of the target soil solution.
Fig 1.
Nonembryogenic citrus cell line of Valencia sweet orange.
Table 3.
The five unique formulations that comprised the experiment.
The experiment was free of ion confounding because only the target ions Na+ and K+ were varied. All other ions in MS medium were held constant.
Table 4.
Recipes generated by ARS-Media for Excel for the 5 media listed in Table 2.
In addition to the salts listed, all recipes included the MS salts that deliver Fe3+, Mn2+, Zn2+, BO33-, I-, Cu2+, MoO42-, and Co2+ as follows: FeSO4.7H2O, Na2EDTA.2H2O, MnSO4.4H2O, ZnSO4.4H2O, H3BO3, KI, CuSO4.5H2O, Na2MoO4.2H2O, and CoCl2.6H2O. These salts did not vary between formulations.
Table 5.
Treatment design points and fresh weight growth data of a 2-component quadratic mixture design.
The proportions of K+ and Na+ were varied while the total amount was kept constant at 23.2 mM, the level in MS medium. For example, treatment point #1 would include 17.4 mM Na+ and 5.8 mM K+.
Table 6.
Recipe generated by ARS-Media for Excel to recreate the soil solution ion composition of Angle et al. [4].
Table 7.
ANOVA of the effect of K+ and Na+ on fresh weight growth of citrus nonembryogenic cells.
The coefficients for K+ and Na+ under the linear mixture are estimates of the response at each vertex, not estimates of the effects of these two ions. The K+ x Na+ term is not an interaction term, though it looks like one, but a quadratic blending term unique to mixture models. This term is used to determine if the mixture components exhibit nonlinear blending and if that blending is synergistic or antagonistic.
Fig 2.
Response plot of the effect of the 2-component K+ and Na+ mixture on % increase in fresh weight of Valencia sweet orange nonembryogenic callus.
The significant linear component, depicted by the dotted line, is the estimated response at each vertex (251% and 1011%), and the significant quadratic component indicates nonlinear blending or that the response deviates (above) from what would be predicted by the linear component. However, the graphic reveals a sharp reduction in fresh weight at 0 mM K+ / 23.2 mM Na+ and essentially a flat relationship between 5.9 mM K+ / 17.3 mM Na+ to 23 mM K+ / 0.2 Na+. This indicates a threshold where Na+ cannot completely substitute for K+.