Cold hardiness testing is an important tool used at US Forest Service nurseries for evaluating the physiological quality of conifer stock. Three tests were compared. A whole-plant freeze test without root system insulation permitted evaluation of cold injury without complication by drought symptoms if plants were misted. A cutting freeze test underestimated expected field foliar injury due to high humidity used to maintain the cuttings. An electrolyte leakage test failed to detect economically important bud mortality when buds were less hardy than other tissues.
Data from all three testing methods were analyzed with an original software package consisting of standardized input files and a step-by-step series of fortran programs and command files for use with commercially available non-linear modelling programs. Injury (y) versus temperature (x) data were modelled with the versatile, 4-parameter Weibull sigmoid model. Injury estimates at specific temperatures (Ix) or temperature estimates causing specific injury levels (LTy) were calculated with confidence and calibration intervals, respectively. The statistical significance of differences between Ix or LTy estimates was then determined.