Plant viability as a function of temperature stress: The Richards function applied to data from freezing tests of growing shoots

Abstract

Frost resistance of growing Salix viminalis L. shoots was determined by rating mortality percentage under two commonly used freezing conditions: a condition in which plants were encased in crushed ice and another in which plants were moistened with tap water prior to freezing. The mortality-temperature data were fitted with a logistic function (having a fixed inflection point halfway between the asymptotes) and with a Richards function, which is a double asymptotic sigmoid function with a variable inflection point. Different frost resistance curves were obtained, depending on the freezing conditions used. However, conditions were inadequate for efficient ice nucleation under either condition. This implies that the applied freezing conditions are not suitable when the purpose is to induce and duplicate early ice crystal formation conditions. The Richards derivatives were negatively skewed in the one case and positively skewed in the other case, giving inflection points, as a function of the upper asymptote, situated at 0.37 when shoots were frosted in the presence of ice and at 0.81 when shoots were frozen in the presence of added moisture. These values differed significantly from 0.50, through which the logistic function would have forced the curves. Because of the significant asymmetry in these frost-resistance curves, the Richards function led to a more accurate reflection of the temperature-mortality course of growing Salix stems than the logistic function. The Richards function possesses the flexibility needed to describe plant injury response in terms of physical and plant physiological mechanisms. Therefore, the Richards function is recommended rather than the logistic function for the assessment of frost resistance.