Table Of ContentsNext Page

Expression of genotypic responses to frost tolerance in wheat

Thusitha A. Gunawardena1, B.P. Naidu1,2 and Shu Fukai1

1School of Land and Food Sciences, The University of Queensland, QLD 4072, Australia. Email
Department of Natural Resources and Mines, Indooroophilly, QLD 4068, Australia.


Yield losses due to frost in Australian wheat crop can be high and are often associated with head-frosting. Two field experiments were conducted over two seasons to investigate the genetic variation in frost tolerance in 150 double haploid lines (DHLs) derived from a cross between Kite and Bindawarra. Glycinebetaine content in the leaf blade during frost acclimation/hardening, cell membrane damage (electrolyte leakage) after frost and grain yield were measured. Significant variation in cell membrane damage was noted (16% to 85%) which was negatively correlated with grain yield (r = - 0.43; p<0.001), indicating genetic variation in frost tolerance among DHLs. Frost acclimation in the first season resulted in large variation in glycinebetaine content (20 to 90 μ mol/g dwt) in the leaf blade which significantly explained the variation in grain yield. It was concluded that genetic variation in frost tolerance among 150 DHLs exists and the results from two seasons showed that the frost acclimation is vital phenomenon in enhancing and expressing genetic variation in frost tolerance.


frost hardening, head-frost


Frost stress during the developmental stages of both male and female reproductive organs reduces wheat yield in parts of Australia by up to 40% in some years. Breeding for long duration cultivars and delaying sowing are main strategies to avoid spring frost. These strategies have a limit as the yield is reduced by 15% per week delay in anthesis past the optimum date (Single 1985). The yield losses associated with delayed sowing or long duration cultivars is primarily attributed to the exposure of grain filling stage of the crop to severe water deficit and high temperatures. While there are several varieties that have been recommended as “frost tolerant” based on the escape mechanism, the search for true genetic tolerance is continuing. Frost acclimation/hardening plays an important role in frost tolerance in current wheat cultivars and differences in frost tolerance of unhardened cultivars are believed to be negligible. Frost resistant cultivars harden faster and deharden more slowly than frost susceptible cultivars (Saulescu and Braun 2001). The investigations reported in this paper attempted to clarify the effects of growth conditions (temperature and soil moisture) prior to frost events in two seasons and the genetic responses.

Plants have evolved various mechanisms to adapt to environmental stresses. One of these mechanisms is the accumulation of low molecular weight organic solutes, which are more commonly known as osmoprotectants. There is increasing evidence that osmoprotectants play protective roles and thus increase tolerance to abiotic stresses. Examples of these solutes include sugars, polyols, amino acids (e.g. proline), their N-methyl derivatives (e.g. betaines) and polyamines (Naidu 1998). Higher glycinebetaine accumulation is associated with frost and low temperature tolerance in wheat (Allard et al., 1998; Naidu et al., 1991), and barley (Nomura et al., 1995). The objectives of this paper are to investigate genotypic variation in frost tolerance among 150 wheat DHLs, and their relationship with the accumulation of glycinebetaine and cell membrane damage, with and without ample frost hardening.

Materials and methods

Two seasons of experiments were conducted during 2002 and 2003 at the DPI - Hermitage research station. In the first season (2002), 150 double haploid lines (DHLs) derived from a cross between Kite and Bindawarra were sown 3 times at 2 week intervals from the 12th April to coincide with frost events (-5C) at various phenological developmental stages. In the second season (2003), seeds were sown 4 times with 2-week intervals starting from 1st April. In each sowing in year 2, 150 DHLs were divided into 3 maturity groups (i.e. early, mid and late maturity) in accordance with their phenological development stage recorded during the first season and planting was staggered by weekly intervals to synchronise head emergence. Field experiments were conducted with 2 m long single row plots with 30 cm inter-row spacing. There were 3 replications in each season. Photoperiod was extended with artificial lighting to midnight to induce and synchronise flowering.

Temperature at the canopy level was continuously recorded by using a ‘TinyTag’data logger. Immediately after each frost event, 3 fully expanded youngest leaves were collected. These leaf samples were cut into 2cm pieces and placed in tubes and 10ml of de-ionised water was added. Samples were shaken for 2 h before tissue electric conductivity (EC1) levels were measured. Leaf samples were frozen with liquid nitrogen after the EC1 measurements and then total electric conductivity (EC2) was measured after the samples were thawed to room temperature. Cell membrane leakage percentage was calculated as: (EC1/EC2 )*100.

One week after the incidence of frost at reproductive stage, foliar damage in the canopy was scored (frost rating) on the scale of 1-5: 1=no damage; 2=10% damage; 3= 25% damage; 4= 50% damage; 5= > 50% damage. When plants were exposed close to sub-zero temperature, prior to frost, leaf samples were collected for the measurement of osmoprotectant (glycinebeataine) accumulation by HPLC method (Naidu, 1998). Genotypes showing re-growth potential after severe frost were noted and grain yield was also recorded at maturity.


Weather condition and phenology

Mean minimum temperature for the 14 day period before any frost event in first season was < - 3.0C, whereas in the second season, plants were exposed to relatively higher temperature (> 2.0C) before frost events (Table 1). In the first season, there were 3 frost events within 4 weeks (30th June to 25th July) period, while only two frost events were recorded in the second season and these events were more than 6 weeks apart. The first season was very frosty with 28 days with minimum temperatures in the range of –5 to –10C (Table 2). Average minimum temperature for the period from the first frost (30th June) to the second (11th July) and, the second to the third (25th July) were –4.5C and –3.2C, respectively, in the first season. Average minimum temperature was 3.3C for the period from the first to the second frost in the second season. Total rainfall recorded for three months period (May to July) was 65.6mm and 136.4mm for first and second seasons, respectively, allowing plants to develop water stress also in the first season. When the second frost occurred in the first season, plants were at three different developmental stages; on average, late booting, heading and flowering stages in late-, mid- and early-sowing, respectively. Staggered planting in the second season aided to achieve near similar developmental stages in all three maturity groups. In the second season, all three maturity groups were near heading and flowering when the frost occurred in early sowing.

Table 1. Frost events recorded in two seasons and average minimum temperature for the period of 14 days before frost.

First season (2002)

Second season (2003)

Date of frost event
(< -5.0C)

Mean minimum temperature (C) for 14 days before frost

Date of frost event
(< -5.0C)

Mean minimum temperature (C) for 14 days before frost

30 June

- 3.2

16 June


11 July

- 4.7

27 July


25 July

- 4.1


Table 2. Occurrence of sub-zero air temperature between May and August in 2002 and 2003 at canopy level.

Air temperature (C)

Number of days



0 to –5



>-5 to –10



< -10






Grain yield

In the first season experiment, there was a significant relationship between cell membrane leakage and grain yield (r = - 0.43; p < 0.01) which was not affected by the reproductive development stage of the DHLs (Figure 1). Cell membrane leakage explained 18% of the variation in grain yield. Cell membrane leakage varied from 16% to 85% while variation in grain yield was 92 to 465 g/m2. In this season, cell membrane

Figure 1. The relationship between cell membrane leakage after the first frost and grain yield (different symbols indicate reproductive development stage of DHLs when the second frost occurred; ●, Late booting to ear emergence; □, Emerged ear to early anthesis; ■, Anthesis). Crossed bars indicate l.s.d. (p=0.05).

leakage in early sowing was highly correlated with the cell membrane leakage in the late sowing (r = 0.82; p < 0.01). Glycinebetaine content in the leaf tissue after frost acclimation (20 to 90 μ mol/g dwt) prior to the incidence of frost was significantly correlated with leaf cell membrane leakage after frost (r = - 0.77; p < 0.00) and also with grain yield (r = - 0.40; p < 0.01).

There was a significant relationship between spike density (spike no/m2) and grain yield, in which spike density explained 49% of the variation in grain yield (r = 0.70; p < 0.01) (Figure 2). In most cases, when the frost rating score exceeded 3.5 (>30% leaf damage) in the scale of 1-5, re-growth of tillers occurred and this assisted to increase grain yield. In other instances, frost rating varied < 3.0 to > 3.5 (severe) but there was no re-growth noted. These results showed that yield variations were associated with the absolute frost tolerance and through recovery or re-tillering. In the second season, staggered planting within each sowing aided to synchronised heading time but exceptionally wet conditions disturbed field establishment in the first and second sowings, leading to complete establishment failure of some DHLs. When frost occurred during heading time in the first sowing all existing DHLs were severely damaged.

Figure 2. The relationship between spike density and grain yield (●, Frost rating scale varied > 3.5 with re-growth; ○, Frost rating scale varied < 3.0 to > 3.5 without re-growth).


A schematic presentation is given in Figure 3 of the association between genetic variation in frost tolerance in wheat under different weather conditions. Although frost occurred around heading stage of wheat DHLs in both seasons, plants in the second season suffered more severe frost damage. When the frost occurred in the second season, after a warmer period, all the existing DHLs suffered severe frost damage, leading to complete yield loss. This was attributed to the lack of frost acclimation in this season. It is believed that the hardening is not a static condition; it changes with temperature, soil and plant moisture, nutrition, and physiological age of plants (Saulescu and Braun 2001). First season results showed that the highest osmoprotectant-accumulating lines suffered almost no frost damage, and vice versa. In this season, limited soil moisture availability and prolonged sub zero temperature prior to frost must have triggered hardening. Therefore, there was a clear evidence of large variation in accumulation of glycinebetaine which led to the maintenance of cell membrane integrity showing less leaf tissue membrane damage.

Figure 3. Schematic presentation of frost tolerance in wheat under two different seasons enhancing genotypic variation.


There is a significant genetic variation for frost tolerance in wheat double haploid lines (DHLs). Sub-zero temperatures prior to frost in the first season enhanced genotypic variation in DHLs. Greater accumulation of glycinebetaine and decreased cell membrane leakage in frost-acclimated DHLs increased frost tolerance and grain yield.


This work was funded by a grant from GRDC.


Allard F, Houde M, Krol M, Ivanov A, Huner NP and Sarhan, F (1998) Betaine improves freezing tolerance in wheat. Plant and Cell Physiology 39, 1194-1202.

Naidu BP (1998) Separation of sugars, polyols, proline analogues and betaines in stressed plant extracts by high performance liquid chromatography and quantification by ultra violet detection. Australian Journal of Plant Physiology 25, 793-800.

Naidu BP, Paleg LG, Aspinall D, Jennings AC, and Jones GP (1991) Amino-Acid and glycinebetaine accumulation in cold-stressed wheat seedlings. Phytochemistry 30, 407-409.

Nomura M, Muramoto Y, Yasuda S, Takabe T, and Kishitani S (1995) The accumulation of glycinebetaine during cold-acclimation in early and late cultivars of barley. Euphytica 83, 247-250.

Saulescu NN and Braun HJ (2001) In “Application of physiology in wheat breeding”.(Eds. MP Reynolds, JI Ortiz-Monasterio, A McNab) Mexico, D. F.:CIMMYT pp 111-123.

Single WV (1985) Journal AIAS 51, 128-134.

Top Of PageNext Page