Electrical resistance changes in different organs of four olive tree (Olea europaea L.) varieties, characterized by different tolerance to chilling and freezing, were examined, during exposure to low temperature. Apparent critical temperatures (CT) and freezing temperatures (Tfr) were identified on the basis of the electrical resistance changes. Both temperatures were lower for the more chilling-tolerant genotypes. From the apparent critical temperatures, the absolute critical temperature (CTabs) and the time delay of the chilling signal transduction process were calculated. In shoots, CTabs varied from 8·8 °C for Ascolana (chilling-tolerant variety) to 13·6 °C for Coratina (chilling-sensitive variety). The magnitude of the transduction time was very similar (about 2 min) for the three genotypes that are more sensitive to chilling, whereas it was significantly higher (about 3 min) for the most tolerant genotype. Different freezing temperatures were observed for different organs. It would appear from this experiment that the order of sensitivity is roots > leaves > shoots > vegetative buds. Accord was found between the absolute critical temperature of electrical resistance and the critical temperature of membrane potential. The occurrence of electrical resistance changes in the tissues of the olive trees exposed to low temperature suggests the use of this experimental procedure as a quick, easy and non-destructive tool to screen plant tissues for chilling tolerance. The strong dependence of the electrical resistance on low temperature, and the critical temperature of around 10 °C, can yield interesting information about the lowest thermal limits for the continuation of normal physiological processes and therefore about the adaptability of plants to particular environments.
Temperature is one of the most important climatic factors conditioning the distribution of plant species. Plants have a thermal range which maximizes growth, and when outside that range their different physiological processes are suspended or slowed down. Plants can survive within thermal limits that are wider than those that allow the normal ontogenetic development. Survival is dependent on the developmental stages and on the type of organ and is also related to the rate at which the minimum temperature is reached and its duration. Thus, for many economically important crops, the specific temperature minimum limits the season of growth, the storage conditions and the geographic distribution, and so breeding for low-temperature-tolerant plants is necessary.
Traditionally, plant screening for chilling and freezing tolerance is based on visual observations in the field. Unfortunately, this screening is subjective and subordinate to other influential factors such as wind, air humidity, exposure, water status and health conditions of the plants that could distort the final result. Moreover, these observations are limited to the aerial part of the plant and they do not consider damage that can occur at the root level. Recently, interesting results to quantify chilling tolerance have come from the study of some antioxidant enzymes ( Wise & Naylor 1987; Kendall & McKersie 1989; Bridger et al. 1994 ; Solecka & Kacperska 1995), and from the chlorophyll fluorescense quenching analysis (CFA) ( Havaux 1987; Ortiz-Lopez et al. 1990 ; Andrew, Fryer & Baker 1995). Although these techniques are effective, they are not sufficiently efficient for breeding programmes ( Shabala & Newman 1997) and are resource and labour-intensive. Therefore, an efficient technique is required, that also allows the evaluation of the injuries caused to the roots by chilling and that, above all, gives rapid and reproducible results.
The olive tree is a warm temperature evergreen plant, which grows mostly between 30° and 45° latitude in both hemispheres. However, during recent years, the increasing demand for olive oil has expanded the cultivation of olive trees into geographical zones at higher latitudes than those of the original Mediterranean basin. Moreover, cool autumns, which slow down the maturation process, improves the quality of olive oil ( Palliotti & Bongi 1996) and this has led to olive trees being cultivated where there is a recurrent danger of frost.
Below − 12 °C the trees suffer severe damage ( Larcher 1970), but also at − 7 °C, damage to the aerial parts of the plant, mainly leaf drop and twig desiccation, can reduce the productivity and threaten the life of the plant ( Palliotti & Bongi 1996). These conditions occur quite frequently in many areas where the olive is cultivated such as in central Italy, where frost has proved lethal on several occasions in the 20th century, and in France, the former Yugoslavia, Azerbaijan, China and USA ( Sakai & Larcher 1987). Consequently, the development of frost-tolerant varieties or clones for this species remains a challenge to plant breeding.
Physiological changes occurring in plants when cooled to freezing temperatures are of great interest in order to understand the mechanism of plant adaptation to low temperature. Since one of the primary mechanisms involved at low temperatures appears to be ionic transport through the cellular membranes, mainly because of the damage to the H+- and K+-transporting systems ( Levitt 1980; Yoshida 1991, 1994; Palta & Weiss 1993; Shabala & Newman 1997), it seems interesting to monitor the state of electrical resistance as a possible indicator of the initial stage of chilling stress in plants.
The present study reports measurements of electrical resistance (R) changes, during exposure to low temperature, in four olive tree varieties characterized by different tolerance to chilling and freezing. The results indicate the possibility to assess the critical and the freezing temperatures within a single experiment. Measurement of such temperatures could be effectively used to differentiate plants for cold tolerance.
MATERIALS AND METHODS
Plant material and growth conditions
One chilling-tolerant (Ascolana), two chilling-sensitive (Coratina and Frantoio), and one intermediate variety (Leccino) of 3-year-old olive trees (Olea europaea L.), grown in 3 L pots containing a 50 : 50 (v/v) mixture of sandy gravel and peat, were used. Plants were grown in a field located in Pescia, Italy (43°54′N, 10°41′E. 30 m asl) and were brought into the laboratory 3 h before the experiment. Samples of current year shoots, leaves removed from the third node from the top, vegetative buds and woody roots, were used for the experiments. All the experiments were conducted in winter on cold-acclimated plants.
Electrical resistance measurements and differential thermal analysis
When a low frequency alternating current is applied to plant tissue it flows through extracellular spaces, its passage through the symplast being limited by the high impedance of the membrane. With increased frequency, the amount of current that passes through the symplast increases as a result of the decrease of membrane impedance ( Cole 1968). Impedance measurements made at low (20 Hz) and high (1 MHz) frequencies therefore reveal information about extra- and intracellular fluids.
The procedure for the electrical resistance measurements was previously described in detail by Mancuso (1999a, b). Briefly, two Ag/AgCl needle electrodes (in order to keep the electrode/tissue interface polarization to a minimum) were inserted in the tissues. The absolute impedance was then measured at 1 MHz using an impedance meter (LX 1192; Nuova Elettronica, Bologna, Italy), with an input voltage level of the sine signal of 20 mV (r.m.s.).
Measurements of electrical resistance changes and differential thermal analysis (DTA) were carried out simultaneously on the same tissue. Excised plant tissues, with inserted electrodes, were placed on one side of a thermopile plate (Melcor CP 1·4–17–10 L; Peltier, USA) and a ‘dried sample of tissue’ on the other used as a reference to detect the differential temperature changes between the two samples during the freezing ( Palliotti & Bongi 1996). A copper–constantan microthermocouple (0·2 mm diameter) was attached to the thermopile plate near the tissue fragment to measure ambient temperature and the whole plate was covered on both sides with a 2-mm-thick layer of closed cell polystyrene tied to the plate with Parafilm to decrease loss of heat to the surroundings. The samples were then placed in a freezing cabinet and cooled at the rates of 0·5, 1·4 and 3·2 °C min−1, usually down to − 30 °C. It had been ascertained in preliminary experiments that Olea europaea tissues did not produce exotherms between − 28 and − 50 °C. Initially temperature was kept at 20 °C, as in the laboratory. Three different plant samples were tested concurrently. The signals from the thermopiles and from the impedance meter, were low-pass filtered, amplified and connected via a multichannel A–D convertor card (Lab-PC-1200; National Instrument, Austin, TX, USA) to a P133 personal computer. Fresh tissues were used for each different rate of freezing.
The absolute critical temperature of electrical resistance (CTabs) and the time delay (τ) in the metabolic processes involved in the cellular response to the changing temperature were calculated according to Shabala & Newman (1997).
Determinations of the lethal freeze temperature by electrical resistance measurements
To test the possibility of using electrical resistance for the determination of the lethal freeze temperature, controlled freezing treatments were carried out in an air-cooled chamber. The initial and final temperature was 20 °C, the rate of cooling and warming 7 °C h−1 and the minimum temperature was maintained for 4 h. At each test temperature, 10 replication of shoot samples (about 1·5 cm long), set in plastic bags, were used. To obtain a reference value for R with all cells dead, some plastic bags and their content were frozen at − 50 °C for 48 h. After thawing the final R was measured and the maximum proportion of R decrement (ΔRmax), was calculated as: ΔR max = Final R/Initial R.
Electrical resistance was measured immediately before (IR) and after (FR) the freezing test and the damage was calculated as the difference in R (ΔR) as a percentage of the ΔR max using the following equation:
Response curves were fitted using the following logistic sigmoid function:
where x = treatment temperature, b = slope at inflection point c, a and d determine the asymptotes of the function.
Membrane potential measurements
The transmembrane potentials (PD) were measured according to a standard electrophysiological technique as described in Rinaldelli & Mancuso (1994).
Tissue samples were horizontally mounted in a 3-mL Plexiglas chamber secured to a microscope stage. Continuously aerated 0·5 mol m−3 CaSO4 solution was permitted to perfuse through the chamber at a flow rate of 10 mL min−1. The outflow, after having passed over the tissue, was collected by aspiration. A copper–constantan microthermocouple (0·2 mm diameter) was placed inside the chamber.
Microelectrodes (tip diameter < 0·5 μm) obtained from single-barrelled borosilicate capillaries (1B F4;, W.P.I., Sarasota, FL, USA) filled with 3 M KCl adjusted to pH 2 to lower the tip potential ( Okada & Inouye 1975) were used. The reference electrode (tip diameter approximately 50 μm) was filled with 3 M KCl in 2% agar (w/v). The Ag/AgCl wire electrodes were connected to a high input impedance (1014Ω) electrometer (homebuilt based on an AD 645 JN operational amplifier). The output signals from the electrometer were low-pass filtered, amplified and connected via the multichannel A–D convertor card to the computer. Measurements were accepted if the electrodes in solution had tip potentials in the range − 5 to − 15 mV and resistance in the range 8–15 MΩ both before and after insertion.
Measurements of PD, to show the effect of temperature, were performed at regular intervals of 1 °C from 2 to 30 °C. At temperature below 2 °C the measurements were not reliable because the solidification of the membrane phospholipids ( Lyons & Raison 1970) makes the adhesion of the membrane to the glass of the microelectrodes imperfect.
The systematic error in the temperature measurements was less than 0·2 °C.
The respiratory activity was determined by measuring the oxygen consumption of a sample, with a fresh weight of 900–1000 mg, inside a chamber containing 25 mL of 0·5 mol m−3 CaSO4. The determination was effected polarographically using a Clark type electrode for pO2 (mod. 0225; E.C.D., Florence, Italy) connected to an oxygen monitor (mod. 8602; E.C.D.), in turn connected via the multichannel A–D convertor card to the computer. The O2 consumption was recorded at intervals of 1 °C from 2 to 30 °C ± 0·2. The temperature inside the chamber was maintained constant by an external water circulation bath. Each measurement was repeated on five different samples at least.
Calculation of activation energy
The rate of repolarization of the cell PD (ΔPD) was measured from the linear portion of repolarization from 5 to 15 °C, with the temperature increasing uniformly with time ( Bravo & Uribe 1981).
For respiration calculations the recorded values at the various temperatures were utilized directly. The trends for the two processes were recorded in logarithmic form on Arrhenius plots as functions of the reciprocal of the absolute temperature. The respective apparent activation energies in kcal/mole were calculated from the slopes of the straight segments of the plots.
All the data derived from the experiments were subjected to ANOVA using the program Statistica version 4·0 (Statsoft, Inc.).
Electrical resistance changes associated with cooling
Ordinarily, the distinctive characters of the electrical resistance kinetic were similar in the different cultivars studied. Figure 1b shows a typical electrical resistance profile for a shoot of olive tree, measured at a rate of 0·5 °C min−1 both for the cooling and the warming phase. It was possible to distinguish five distinct phases in response to the variation of temperature:
1 Initially the electrical resistance was stable. This phase finished with the beginning of the electrical resistance increase, at a critical temperature, designated as CT in the graph. The value of CT was determined graphically as showed in Fig. 1b.
2 The electrical resistance continued to increase as the temperature was reduced to freezing (Tfr in the graph). Figure 1a shows the simultaneous recording of the exotherm measured by DTA.
3 After freezing, within a few seconds the electrical resistance reached its maximum and then remained almost stable with the subsequent increase of the temperature.
4 With the thawing, the electrical resistance decreased to reach a stable minimum that was less than the initial R.
5 Subsequent increases of the temperature did not have any effect on the electrical resistance.
The multiphase pattern described above was remarkably similar in the different genotypes tested although the values of CT, as well as those of Tfr varied greatly ( Fig. 2). Table 1 shows the mean values of CT and Tfr for different organs of the four cultivars studied. Clear differences appear for the values of CT and Tfr. Both CT and Tfr were higher for the chilling-sensitive cultivars (Coratina and Frantoio) in comparison with chilling-tolerant Ascolana. Moreover, such differences remained in all the organs studied with the exception of the roots which, from this point of view, are the least meaningful, giving values of CT and Tfr that were very similar among the different cultivars.
Table 1. Apparent critical temperatures and freezing temperatures associated with electrical resistance changes as related to plant chilling and freezing sensitivity. Rate of cooling is 0·5 °C min−1. The results are reported as means ± SE. Values in the row followed by differing letters, significantly differ at P < 0·05
−9·1 ± 0·3 (n = 12)a
−8·8 ± 0·3 (n = 12)a
−9·0 ± 0·3 (n = 12)a
−8·6 ± 0·3 (n = 12)a
8·5 ± 0·2 (n = 12)c
10·3 ± 0·3 (n = 12)b
10·6 ± 0·3 (n = 12)ab
11·2 ± 0·4 (n = 12)a
−14·5 ± 0·6 (n = 12)c
−12·9 ± 0·5 (n = 12)b
−12·3 ± 0·5 (n = 12)ab
−11·8 ± 0·4 (n = 12)a
7·4 ± 0·3 (n = 10)c
9·9 ± 0·4 (n = 11)b
11·5 ± 0·4 (n = 12)a
12·3 ± 0·5 (n = 12)a
−18·6 ± 0·8 (n = 10)c
−15·2 ± 0·7 (n = 11)b
−13·2 ± 0·7 (n = 12)a
−12·6 ± 0·6 (n = 12)a
7·6 ± 0·3 (n = 9)c
8·9 ± 0·3 (n = 10)b
8·8 ± 0·4 (n = 10)b
10·2 ± 0·4 (n = 10)a
−19·3 ± 1·1 (n = 9)c
−16·2 ± 0·8 (n = 10)b
−17·3 ± 1·2 (n = 10)b
−13·5 ± 0·8 (n = 10)a
8·9 ± 0·4 (n = 12)b
9·3 ± 0·3 (n = 12)ab
9·6 ± 0·4 (n = 12)ab
9·8 ± 0·4 (n = 12)a
From the air-cooled chamber tests, a clear increase in ΔR with decreasing minimum temperature was observed in the shoots of the olive plants ( Fig. 3). The relationship, as expected, was sigmoidal, and the inflection point, which predicts the lethal freeze temperature (LT50; lethal temperature at which 50% of the tissues are injured) was significantly different between sensitive and tolerant cultivars, and in good agreement with the Tfr detected for shoots in Table 1.
Critical temperature at different rates of temperature change
Critical temperatures of the shoots during cooling were recorded at different rates of temperature change (Tr) to test the existence of time delay in the metabolic processes involved in the cellular response to the changing temperature. Figure 4 represents the effect of different rates of cooling on the CT detected by electrical resistance changes as in Fig. 1. At all rates the olive cultivars with greater chilling-tolerance showed lower values of CT. In some cases the differences in the CT at the different cooling rates were dramatic. In Ascolana, a chilling-tolerant genotype, for example, CT varied from − 0·7 °C for Tr = 3·2 °C min−1 to 7·2 °C for Tr = 0·5 °C min−1. Coratina, the more chilling-sensitive cultivar that is a native of southern Italy (Apulia) and usually cultivated in areas with high average temperatures, showed the highest values of CT at every Tr (7·0–12·4 °C). The cultivar Leccino with moderate chilling-tolerance had intermediate values of CT.
For the estimation of the absolute critical temperature (CTabs) of electrical resistance, a correction factor must be taken in account. Thus, CTabs for the R of the shoots was calculated using a linear regression between the rate of temperature changed utilized and the CT values obtained. Figure 5 shows that the relationship is about linear over the range of Tr studied. The CTabs is the intercept of the line of best fit on the Y-axis at Tr = 0. Absolute critical temperature ranged from 8·8 °C, for the chilling-tolerant Ascolana, to 13·6 °C for the chilling-sensitive Coratina ( Table 2). The magnitude of the time constant (τ) (representing the time delay of signal transduction) also varied, being greater for the most chilling-tolerant genotype (Ascolana, 2·90 min) in comparison with the less tolerant genotypes (from 1·94 to 2·01 min).
Table 2. Absolute critical temperatures in shoots of olive tree. The results reported as means ± SE were calculated from n = 12 experiments in 12 different plants. Values in the column followed by differing letters, significantly differ at P < 0·05
13·6 ± 0·4a
2·0 ± 0·1b
8·8 ± 0·3d
2·9 ± 0·2a
10·6 ± 0·4c
2·0 ± 0·2b
12·6 ± 0·3b
1·9 ± 0·2b
Effects of low temperature on membrane potential and respiration
Cell PD was observed to decrease with decreasing temperature in the cortical cells of the shoots of the cultivar Leccino chosen for its intermediate sensitivity to the low temperatures ( Fig. 6).
Membrane potential and respiration show different responses to temperature ( Fig. 6). The Arrhenius plots of the rate of repolarization of the cell PD and of the respiration rate present sharp variations of apparent activation energy at different critical temperatures (respiration 30·6 kcal/mole versus 24·4 kcal/mole at 14 °C; PD 53·2 kcal/mole versus 25·3 kcal/mole at 11 °C) ( Fig. 7). Above 15 °C the PD was much less dependent on temperature and therefore reliable information can not be extracted.
The validity of the performance of an electrical resistance measuring system in the detection of changes in R against an artificially generated cool gradient was initially established. Electrical resistance showed multiphase kinetics directly linked to the changes in the temperature. The most important result of this study was that CT and Tfr detected from R changes were strongly correlated with the chilling-tolerance of different varieties of olive trees, suggesting a good predictive power of the R measurements for screening in plant breeding purposes.
Electrical resistance measurements allowed an easy detection of freezing temperatures as shown by the concurrent DTA analysis. Contrary to other studies ( Ishikawa 1984), only one type of exotherm was observed: the low temperature exotherm produced by the freezing of the intracellular solution. The high temperature exotherm, which can indicate freezing of vascular and intercellular water, was not detected, probably because free water accounted for a minimum of total water. The magnitude of Tfr is representative of the thermal tolerance limits only in deep supercooling plant types which are subject to a freeze-avoidance mechanism linked to water tensioactivity ( Bowers 1994). In this study Tfr, revealed by R changes and by DTA, correlated closely with the LT50 calculated from the ΔR measurements confirming both the reliability of ΔR measurements for the determination of the lethal freezing temperature and the occurrence of deep supercooling in Olea europaea as already suggested ( Larcher 1970).
Different Tfr in different organs was observed. It would appear from this experiment that the order of sensitivity is roots > leaves > shoots > vegetative buds. It is also of interest that the roots were the only tissues that did not show differences between susceptible and tolerant genotypes. This could be significant in terms of evolutionary dynamics since it is exceptionally rare that the ground temperature is less than − 9 °C in the geographic areas where the olive is mostly cultivated.
Critical temperatures at different rates of cooling temperature were not the same, showing that some time is necessary for the transduction of the temperature signal from the initial sensing receptor to the effector action that results in the changes in R. Because of this transduction delay, the error in the critical temperature measurement (mainly at the higher cooling rate) could amount to several degrees. A simple analytical procedure proposed by Shabala (1996) allowed calculation of the absolute critical temperature providing, moreover, another informative indicator of plant adaptability to low temperature: the transduction time (τ). The magnitude of τ was very similar (about 2 min) for the three genotypes that were more sensitive to chilling, whereas it was significantly higher (about 3 min) for the most tolerant genotype. These results correspond with temporal constants of signal transduction for other environmental changes. For example Shabala & Newman (1997) showed time constants of about 3 min for the activation/deactivation of the H+ influx in response to low-temperature, in maize.
Agreement was found between the value of the CTabs of R and the critical temperature detected from the Arrhenius plot relative to the effects of temperature on cell PD for the cortical cells in shoots of the cultivar Leccino. Recently, a critical temperature of around 9 °C was reported for active and passive H+ transporter in corn roots ( Shabala & Newman 1997). These findings support the hypothesis that measurements of R show directly the effect of temperature on all of the factors responsible for the membrane potential. The results from the Arrhenius plots relative to the effects of temperature on cell PD and respiration, indicate critical temperatures of 11 and 14 °C, respectively, in shoot tissues of the cultivar Leccino. The transition temperature at 11 °C is in good agreement both with the values of CTabs in the range 8·5–13·5 °C, and with the temperature of lipid phase transition reported by Lyons (1973). As the primary event in chilling injury is the alteration in the state of cellular membranes from a relatively fluid liquid-crystalline state to a less fluid gel state ( Lyons & Raison 1970), the different physiological responses to low temperature in the different genotypes studied, could be attributed to the physical state of the membrane determined by its lipidic composition ( Lyons & Raison 1970), in particular by the ratio between saturated and unsaturated fatty acids. With the prevalence of unsaturated fatty acids there would be a lowering of the critical temperature, under which processes of solidification of the membrane’s phospholipids begin ( Lyons & Asmundson 1965). A higher content of unsaturated fatty acids would therefore cause an increase in chilling-tolerance ( Murata et al. 1992 ).
In conclusion, the measurement of the electrical resistance changes in plants exposed to low temperature appears to be an useful index for estimating chilling sensitivity in plant breeding programmes and provides a quick, easy and non-destructive experimental procedure.