Standardization of electrolyte leakage data and a novel liquid nitrogen control improve measurements of cold hardiness in woody tissue

Abstract

Background: A variety of basic and applied research programs in plant biology require the accurate and reliable determination of plant tissue cold hardiness. Over the past 50 years, the electrolyte leakage method has emerged as a popular and practical method for quantifying the amount of damage inflicted on plant tissue by exposure to freezing temperatures. Numerous approaches for carrying out this method and analyzing the resultant data have emerged. These include multiple systems for standardizing and modeling raw electrolyte leakage data and multiple protocols for boiling or autoclaving samples in order to maximize leakage as a positive control. We compare four different routines for standardization of leakage data and assess a novel control method-immersion in liquid nitrogen in lieu of traditional autoclaving-and apply them to woody twigs collected from 12 maple (Acer) species in early spring. We compare leakage data from these samples using each of four previously published forms of data analysis and autoclaving vs. liquid nitrogen controls and validate each of these approaches against visual estimates of freezing damage and differential thermal analysis.

Results: Through presentation of our own data and re-analysis of previously published findings, we show that standardization of raw data against estimates of both minimum and maximum attainable freezing damage allows for reliable estimation of cold hardiness at the species level and across studies in diverse systems. Furthermore, use of our novel liquid nitrogen control produces data commensurate across studies and enhances the consistency and realism of the electrolyte leakage method, especially for very cold hardy samples.

Conclusion: Future leakage studies that relativize data against minimum and maximum leakage and that employ our updated liquid nitrogen control will contribute generalizable, repeatable, and realistic data to the existing body of cold hardiness research in woody plants. Data from studies conducted using a liquid nitrogen (and not an autoclaving) control can still be compared to previously published data, especially when raw data are standardized using the best-performing approach among those we assessed. Electrolyte leakage of woody twigs emerges as a useful technique for quickly assessing the probability of tissue death in response to freezing in dormant plants. Differential thermal analysis may provide different and complementary information on cold hardiness.

Keywords: Acer; Cold hardiness; Differential thermal analysis; Electrolyte leakage; Freezing tolerance; Maple.

Fig. 1

Schematic illustration comparing different approaches to measuring electrolyte leakage. A In the Anderson approach (no minimum or maximum leakage specified, yields R), samples may not reach an “Absolute 50%” damage point. Instead, the temperature at which “Relative 50%” damage is attained may be more meaningful. B By comparison, values of I, in the Flint approach (orange line) are zeroed, although this may not drastically displace the curve relative to an Anderson curve. C In the Lim approach, data are stretched between 0 and 100% damage, usually at the warmest and coldest temperature treatments, respectively. D Use of a liquid nitrogen control is expected to elevate all leakage values, making, for instance, an Anderson curve behave more like a Lim curve (“Absolute” and “Relative” 50% values similar), and improving generalizability among approaches

Fig. 2

Comparison of four approaches for fitting curves to data representing the relationship between freezing damage and temperature in (A, C, E, G) A. caudatifolium and (B, D, F, H) A. campestre stem segments (plots for other species provided in Additional file 1). Curves fit to data on a per-genotype (red, blue, and green) and per-species (black curve) basis are fit in each case. Panels show curves fit following the approach of A, B Anderson et al. [12], C, D Flint et al.[7], and Lim et al. [13]. Approaches vary, as indicated, in their use of room-temperature (zeroing; C–H) and deep freezing (maximum damage; E–H) controls and reliance on general logistic (A–F) vs. Gompertz (G–H) curves

Fig. 3

Damage, as reflected by electrolyte leakage (solid lines) and visual estimates (dashed lines), induced by freezing from − 10 to 80 ℃. Electrolyte leakage is calculated using the Lim_logistic approach. Panels represent estimates of damage to particular genotypes. Color-coding indicates species

Fig. 4

Visual cambial damage corresponded to critical cold hardiness estimated from electrolyte leakage data. Values of T50 given here (Table 1) are calculated using the Lim_logistic approach. Representative stem samples following freezing are shown for A Acer caudatifolium, B A. davidii, C A. hyrcanum, and D A. negundo. Scale bar = 0.5 cm

Fig. 5

Fitness characteristics of the relationship between fitted values of electrolyte leakage and visual damage. Red contour delimits the area where: A Correlation is greater than 0.55; B Bias < abs (5 ℃); and C RMSE < 7 ℃. Dashed rectangle delimits data used in Additional file 4

Fig. 6

A Sample conductivity following boiling predicts conductivity following immersion in liquid nitrogen across a range of values and for diverse species (color-coding indicates species as in Fig. 2). Circled points are statistical outliers and lines indicate Deming regression error. r² were calculated based on residuals in each direction. B When a boiling standard is used, electrolyte leakage values derived using different curve-fitting procedures (e.g. Anderson vs. Lim_logistic) are not comparable above ~ 25% leakage. C However, use of a liquid nitrogen standard makes outputs of these two routines more comparable. Grey bar indicates a range of values within 15% of the 1:1 line