Changes in cold tolerance and the mechanisms of acclimation of photosystem II to cold hardening generated by anther culture of Festuca pratensis × Lolium multiflorum cultivars


  • Marcin Rapacz,

    Corresponding author
    1. Agricultural University of Cracow, Faculty of Agriculture and Economics, Department of Plant Physiology, Podluzna 3, 30–239 Kraków, Poland;
      Author for correspondence: Marcin Rapacz Tel: +4812 425 33 01 Fax: +4812 425 33 20 Email:
    Search for more papers by this author
  • Dagmara Gasior,

    1. Agricultural University of Cracow, Faculty of Agriculture and Economics, Department of Plant Physiology, Podluzna 3, 30–239 Kraków, Poland;
    Search for more papers by this author
  • Zbigniew Zwierzykowski,

    1. Institute of Plant Genetics, Polish Academy of Sciences, Strzeszyñska 34, 60–479 Poznañ, Poland;
    Search for more papers by this author
  • Agnieszka Lesniewska-Bocianowska,

    1. Institute of Plant Genetics, Polish Academy of Sciences, Strzeszyñska 34, 60–479 Poznañ, Poland;
    Search for more papers by this author
  • Michael W. Humphreys,

    1. Institute of Grassland and Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion, SY23 3EB, UK
    Search for more papers by this author
  • Alan P. Gay

    1. Institute of Grassland and Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion, SY23 3EB, UK
    Search for more papers by this author

Author for correspondence: Marcin Rapacz Tel: +4812 425 33 01 Fax: +4812 425 33 20 Email:


  • • Mechanisms of photosynthetic acclimation to cold were investigated on androgenic plants generated from Festuca pratensis × Lolium multiflorum (4x) cultivars Felopa and Sulino and on parental material.
  • • Photosynthetic acclimation and resistance to high-light induced inactivation of PSII at low temperature were studied using chlorophyll fluorescence techniques in relation to winter hardiness, frost resistance and cold acclimation in field and controlled conditions.
  • • In the field increased energy dissipation before winter through a lower maximum quantum yield of PSII was correlated with improved winter survival of these genotypes. In controlled conditions winter hardy plants were more resistant to cold-induced inactivation of PSII. During cold acclimation of winter hardy plants nonphotochemical quenching (NPQ) increased, except in one genotype where photochemical quenching increased.
  • • The use of androgenic lines revealed gene combinations that determined alternative photoinhibition avoidance mechanisms in the parental genome. Increased dissipation of light energy is an alternative process to the increased photosynthetic capacity reported previously to be the main mechanism of photosynthetic acclimation to cold in herbaceous Poaceae.

fluorescence of leaves in the dark when all PSII reaction centres are open

inline image

fluorescence in leaves previously exposed to light darkened just before measurement


variable fluorescence (Fv = Fm − F0)

Fm and inline image

fluorescence when all PSII reaction centres are closed in dark- and light-exposed leaves, respectively


steady state fluorescence in light exposed leaves

Fv : Fm

maximum quantum yield of PSII


photosynthetic photon flux density


high-light treatment


temperature causing a 50% reduction in re-growth rate after freezing


photosystem II


current quantum yield of PSII


nonphotochemical quenching of chlorophyll a fluorescence


photochemical quenching of chlorophyll a fluorescence

HF and HS

followed by a number are the androgenic genotypes derived from F. pratensis ×L. multiflorum (4x) (Festulolium) cultivars Felopa (winter hardy) and Sulino (winter susceptible), described in Table 1

Table 1.  Origins of androgenic plants generated from Festuca pratensis × Lolium multiflorum (4x) cultivars Felopa and Sulino in the experiments. In this outbreeding material each seed sown from the more hardy (Felopa) and less hardy (Sulino) varieties has a different genotype, identified by a number below. The label H represents an androgenic plant derived from Felopa (F) or Sulino (S), with an identifying number
Initial genotypeNo. of androgenic plants obtainedNo. of plants grown in the field scored for winter hardinessPlants used only for field chlorophyll fluorescence measurementsPlants used for laboratory and field chlorophyll fluorescence measurements
Felopa 16 12 5HF-40 
Felopa 18 7532HF-58, HF-100, HF-106Felopa 18, HF-118, HF-105, HF-112, HF-94
Felopa 20 18 8HF-125, HF-169 
Sulino 4 24 9 Sulino 4
Sulino 5 14 3HS-1126 
Sulino 7 2174HS-436, HS-529 
Sulino 1610760HS-196, HS-210, HS-211HS-197, HS-212, HS-262


Recently anther culture procedures have been developed to generate large populations of diverse genotypes from Lolium× Festuca F1 hybrids (Humphreys et al., 1998; Zwierzykowski et al., 1999; Zare et al., 2002) and Festulolium cultivars (Lesniewska et al., 2001). The diversity arises because gametes from out-breeding Lolium and Festuca species are highly heterogeneous and each has a unique gene combination. Androgenesis reduces masking by dominant alleles and has produced genotypes rarely recovered by conventional breeding programmes (Humphreys et al., 1998). It may also initiate desirable epigenetic and pleiotropic effects (Humphreys et al., 2003). Some of the androgenic genotypes produced have extended the range of useful traits over that in parental material, and here we are particularly interested in the increased range of cold tolerance (Zare et al., 1999).

An important aspect of cold tolerance is the avoidance of photoinhibition of photosynthesis that can occur during periods of relatively high light and low temperature in autumn and winter. Photoinhibition can occur if the rate of light harvesting by PSII exceeds the capacity of electron transport, which is reduced by low temperatures. Resistance to photoinhibition is related to cold tolerance (Huner et al., 1998; Pocock et al., 2001) and changes in the redox state of PSII have been proposed as a temperature sensing mechanism for cold acclimation (Huner et al., 1996, 1998; Ndong et al., 2001) and de-acclimation (Rapacz, 2002a,b). Many alternative mechanisms for dissipation of excess energy harvested by PSII, which reduce the potential for photoinhibition have been demonstrated (Huner et al., 1998; Adams et al., 2002). However, as cold-acclimation of cold tolerant herbaceous plants requires energy, they must maintain photosynthesis as well as avoiding photoinhibition (Huner et al., 1993; Rapacz & Janowiak, 1998). Thus in view of the changes in cold tolerance reported for the androgenic genotypes and the possibility that new mechanisms may be revealed by the androgenesis, their photosynthetic characteristics, with reference to avoidance of photoinhibition will be studied here.

Materials and Methods

Plant materials and field experiment

Eighteen genotypes of each of two Polish Festulolium cultivars, Felopa (more winter hardy) and Sulino (less winter hardy) derived from F. pratensis × L. multiflorum amphiploid (2n = 4x = 28) F8 hybrids (Zwierzykowski et al., 1998) and 263 from 882 androgenic genotypes derived from both cultivars (Leœniewska et al., 2001), were grown over three winters (1999/2000, 2000/2001 and 2001/2002) in a field experiment. Three replicate clones of each genotype were planted as spaced plants with centres on a 0.45-m grid on 23 August 1999 in fully randomised design in a field plot at Lopuszna, Poland (20°08′-E, 49°28′-N, altitude 568 m) on a brown soil. Clones were cut back four times during the growing season and weeds between plants were removed before each cut. Before trial establishment 50 kg N, 60 kg K2O, 40 kg P2O5 ha−1 fertilizer was applied, then 30 kg N ha−1 after each cut and 20 kg N, 70 kg K2O, 60 kg P2O5 ha−1 after the last cut in autumn. Maximum and minimum temperatures during each 24-h period were recorded using radiation shielded probes (maximum) and a toluene-filled thermometer (minimum) 1 m above the ground, at a distance of 200 m from the experimental site (Fig. 1).

Figure 1.

Minimum (solid) and maximum (dashed) temperatures in the field in Lopuszna during the winter seasons of 1999/2000, 2000/2001 and 2001/2002.

In each of the three winters winter hardiness was estimated as detailed below. Also current quantum yield of PSII (φPSII) and maximum quantum yield of PSII (Fv : Fm) were measured as described below in the field on 10 October 2000 and 20 October 2001 on 19 androgenic genotypes (origins listed in Table 1) chosen to give a range of winter survival. Felopa 18 was also measured as the parent of the majority of HF-lines studied. Sulino 16 and Sulino 7, the parents of the majority of the HS genotypes studied were killed during the first winter so a genotype with similar winter survival (Sulino 4) was measured instead. Measurements were made on 10 leaves over the cloned replicates, at 13°C ± 1.1 and 10°C ± 0.8 and natural field irradiance 200 ± 15 and 150 ± 18 µmol m−2 s−1 in 2000 and 2001, respectively.

Winter hardiness estimation

Overall winter hardiness was split into two components, winter survival and spring regrowth. Winter survival was scored visually using a scale of 0–9, where the extremes were 0, plants without green tillers and 9, plants without any visible winter damage. Scores were recorded on 4 April 2000, 15 April 2001 and 1 March 2002 on clones maintained continuously in the field. Spring regrowth after cutting on the day of winter damage estimation was also scored visually by estimating the size of regrowing plant using a visual score from 0 (no regrowth) to 9 (maximum regrowth) on 10 May 2000, 30 May 2001 and 18 April 2002. In this environment some regrowth after cutting did occur from basal tillers on plants that at the time of scoring winter survival had no live tillers visible.

Measurements of chlorophyll fluorescence

All PSII chlorophyll a fluorescence measurements were made with a FMS2 fluorometer (Hansatech Ltd. Kings Lynn, UK) on the middle section of the youngest, fully expanded leaves. Before Fv : Fm measurements, leaves were dark adapted for at least 15 min (30 min in the field) in leaf-clips and values of inline image and Fs were recorded when Fs became stable after re-exposure to light. Then inline image was measured on removal of the light but with the addition of far-red light to ensure rapid opening of PSII reaction centres. Current quantum yield of PSII was calculated according to Genty et al. (1989): φPSII = (inline image − Fs)/inline image. Photochemical quenching coefficient (qP) was calculated according to Schreiber et al. (1994): qP = (inline image − Fs)/(inline image − inline image). Nonphotochemical quenching (NPQ) was calculated as: NPQ = (Fm − inline image)/inline image (Bilger & Bjorkman, 1991).

Laboratory studies of photosynthetic acclimation to cold and frost resistance

Cold acclimation  On the basis of field winter survival in 1999/00 and 2000/01 plants (Table 1) with repeatable good and poor winter hardiness were transferred on 11 September 2001 from the field. The clones were planted in 1.5 dm3 pots and grown in an unheated glasshouse (5–10°C, average 8°C) between November and the end of March and open-air vegetation room for the rest of the year. During the summer season plants were cut every 3 wk, fertilized after each cut with 20 g of multicomponent lawn fertilizer (Kronen, Poland) 20.8% N, 7% P2O5, 11.3% K2O, Mg, Ca, micronutrients. On 12 May 2002 plants were divided into 10, two or three tiller plants per genotype, each placed in 20 cm pots containing sand : peat (1 : 1) (for studies of chlorophyll fluorescence and accumulation of anthocyanins) and into 40 single-tiller plants in 20 cm3 plastic test tubes containing the same soil mixture (for freezing tests). For 2 wk after division plants were grown in an air-conditioned glasshouse (+21/+17°C, day/night, natural photoperiod, about 12/12 h). Then plants were moved to a growth chamber for prehardening (1 wk at +12°C, photoperiod 10/14 h, 300 µmol m−2 s−1 PPFD, Philips AGRO sodium light source, Philips Lightning N.V., Turnhout, Belgium) and then cold acclimation (3 wk at +2°C, with the same PPFD and photoperiod). The experiment reported here was a repeat of one carried out in Autumn 2001 directly after plants were transferred from the field giving similar results.

Measurements of chlorophyll fluorescence and anthocyanin content  Chlorophyll fluorescence was measured 1 d after the start of prehardening (termed as measurements before cold acclimation) and after cold acclimation on the last day at 2°C. All measurements were done at leaf temperature 5°C ± 0.8 in 10 replicates per genotype. Fs was measured at 500 µmol m−2 s−1 PPFD using the actinic light source of the FMS2. The changes in Fv : Fm, φPSII, qP and NPQ caused by cold hardening were expressed as their value after cold hardening divided by their value before cold hardening. Genotype means were calculated from randomly paired measurements before and after acclimation on replicates from each genotype. Because statistical distribution of ratios from values that are also ratios are unlikely to be normal, bootstrap techniques (Efron & Tibshirani, 1993) were used (applying the bootstrap procedure of Genstat 6.1, third edition, Payne, 2002) to calculate 95% confidence limits of the means. The resulting confidence limits, many of which were asymmetrical, with different directions of asymmetry between genotypes are shown in Fig. 2.

Figure 2.

Relationship between winter survival in spring 2001 (a) and 2002 (b) and apparent quantum yield of PSII (Fv : Fm, closed circles) and current photochemical efficiency (φPSII, open circles) measured on October 10, 2000 (a) and October 20, 2001 (b) under field conditions. Data for HS-262 are marked with asterisks. Regression lines were fitted between winter survival and Fv : Fm (correlation coefficients significant P < 0.01).

Leaf anthocyanin content was estimated at the end of cold acclimation using a visual score (0–3), where: 0, no visible anthocyanins; 1, anthocyanins visible along main veins; 2, spots of anthocyanins presents on the leaf blade; 3, leaves evenly coloured with anthocyanins. Scores were assessed on the youngest, fully expanded leaf on each tiller in the clone (20–40) tillers and averaged.

Frost resistance estimation

Freezing tolerance of plants was estimated by the modified method of Larsen (1978). Cold-acclimated plants in test tubes were transferred for 2 d to a growth room at 0/−2°C (day/night), 80 µmol m−2 s−1 PPFD, 10/14 photoperiod. Before the second night some pieces of ice were added to the tube to avoid supercooling and ensure proper freezing. After the second night the temperature was decreased at a rate of 4°C h−1 to −8°C in the dark and after 2 h 10 plants were returned to the conditions of cold-acclimation (+2°C, photoperiod 10/14 h, 300 µmol m−2 s−1 PPFD). The temperature of the rest of the plants was then successively reduced to −10, −12, −14 and −16°C at a rate of cooling of 4°C h−1 and 10 further plants returned to the conditions of cold acclimation after 2 h at each temperature. Twenty-four hours after the last plants had been removed all were transferred to +12°C, 12-h photoperiod, 300 µmol m−2 s−1 PPFD. Control plants were cold hardened at 2°C as for the experimental plants, but not frozen. They were then transferred to 12°C at the same time as plants subjected to freezing. All plants were cut on the day of transfer to 12°C and the regrowth was estimated after 1 and 3 wk by Larsen's visual score (1978): 0, completely dead plant, no sign of leaf elongation; 1, dead plant, but elongation c. 5 mm has occurred; 2, dead plants, but the leaves have elongated 1–2 cm before dying; 3, the plant is dying, but the leaves have elongated to 2 cm or more; 4, the plant may die. It has been growing, but inner leaves may be brown; 5, the plant may survive, but is badly damaged, the regrowth is often discoloured and curled; 6, plant is surviving, but shows severe damage to c. 50% of the leaves; 7, plant is alive, but symptoms of freezing injury are seen (some leaves discoloured or deformed); 8, only tops of some inner leaves discoloured or deformed; 9, no symptoms of injury. The temperature causing a 50% depression in regrowth score compared with control plants (RT50) was estimated from a linear regression fitted to the central part of the sigmoid relationship between the freezing temperature and regrowth score using at least three temperatures.

Responses of chlorophyll fluorescence in hardened and unhardened plants exposed to high light

The experiments were performed in November/December 2002 on two winter hardy genotypes (HS-262, HF-105) contrasting in photosynthetic response to cold acclimation and two winter-susceptible genotypes (HS-436, HS-529). Plants were transferred on 31 October 2002 from an open-air vegetation room to an air-conditioned glasshouse (20/17°C, day/night, 12/12 h photoperiod maintained by Philips Agro sodium lamps, 200 µmol m−2 s−1). After 3 or 6 wk plants were transferred to the conditions of prehardening and cold-acclimation (as already described). Measurements were made on both cold acclimated and nonacclimated plants (transferred directly from the glasshouse) at +2°C. The initial measurements were made in dark-adapted plants subsequently exposed to high light (HL, 1200 µmol m−2 s−1 halogen light source: Decostar 51S 50 W, Osram, Germany) for estimation of Fs, inline image and inline image; these were re-measured after 2, 4 and 6 h of HL. Fm and F0 were measured after 15 min of dark adaptation after measurements of Fs during HL treatment. Measurements were made in two independent series (on different clones of each genotype and at 3-wk intervals), five replicates each. The same leaves were measured before and during HL and results are the mean of both experiments.


Relations of chlorophyll fluorescence parameters to cold hardiness in the field

During winter 2000/01 there was a highly significant (P < 0.001) correlation between Fv : Fm in the autumn, and subsequent winter survival and spring regrowth (Table 2). Plants with high winter survival and spring regrowth had lower Fv : Fm in the autumn (Fig. 3). There was no significant relationship of current photochemical efficiency of PSII with winter survival (Table 2) or of Fv : Fm with φPSII (r = 0.175, P > 0.435). There was also a significant negative correlation of winter survival with steady state fluorescence in light adapted leaves (Fs, Table 2). Other chlorophyll fluorescence parameters measured in autumn did not correlate with winter hardiness (Table 2).

Table 2.  Correlation coefficients (Pearson’s, two-sided) between genotype means of winter hardiness in winter 2000/01 and genotype means of field measurements of chlorophyll fluorescence in autumn 2000
 Fsinline imageφPSIIF0FmFvFv : Fm
  1. Correlation significance: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Winter survival−0.550**−0.510*−0.188−0.075−0.311−0.350−0.730***
Spring regrowth−0.150 0.000−0.117 0.121−0.103−0.144−0.665**
Figure 3.

Relative change in chlorophyll fluorescence parameters during cold acclimation of hardy and nonhardy genotypes in Table 3. The vertical lines at the top of the histogram bars are the 95% confidence limits calculated using bootstrap techniques.

During the second winter (2001/02) the correlation between Fv : Fm in autumn and winter survival was also significant (r = −0.669, P < 0.01). The other parameters of chlorophyll fluorescence in Table 2 were not correlated with winter survival. The range of winter survival was much lower than in the previous year (Fig. 2). This was probably a consequence of winter starting early (Fig. 1) and particularly of lower temperatures from the beginning of November to two-thirds of the way through December than during the same period in the previous year. The sums of accumulated day-degrees below 0 in this period were −180.3 and −668.1 for 2000/1 and 2001/2, respectively. The values of Fv : Fm were also lower than in the previous year. This may be related to the lower temperatures before measurement in 2001 than in 2000 (Fig. 1), which in this material would lead to increased cold hardening which would give lower Fv : Fm (Fig. 2). However, the correlation coefficients between winter survival and Fv : Fm for the 2 yr were not significantly different (P > 0.05 using Fishers z transformation and a test for equality of correlation coefficients, Zar, 1996). This suggests that relative values of Fv : Fm measured in autumn may be a predictor of winter survival irrespective of the conditions experienced in the following winter and the average values of Fv : Fm.

Effects of cold hardening under controlled conditions on chlorophyll fluorescence

Genotypes were grouped according to their descending winter hardiness and frost resistance (Table 3). The main criterion of grouping was the laboratory test of frost resistance verified with values of winter survival and spring regrowth. There was generally a consistent relationship between the temperature required to reduce regrowth by 50% (RT50) measured under controlled conditions and field assessments of winter survival and regrowth. Cold hardening generally reduced Fv : Fm more in plants with greater winter hardiness than in plants with less winter hardiness, although there was some overlap in the intermediate ranges (Fig. 2).

Table 3.  Anthocyanin content and freezing tolerance after cold acclimation under controlled conditions and winter hardiness in the field of genotypes of varying degrees of winter hardiness. Anthocyan content was assessed on a scale of 0–3, and winter hardiness scores were assessed on a scale of 0–9 as described in the text
GenotypeAnthocyanin content scale 0–3Winter hardiness scores scale 0–9Temperature causing 50% depression of regrowth in freezing test RT50 (°C)
 Winter survivalSpring regrowth 
Winter hardy plants
Felopa 18015.558−15.4 ± 0.8
HF-11800.661.6741.5−15.9 ± 0.7
HS-26210.66448.5−15.6 ± 0.6
HF-10521.3344.667.5−15.8 ± 0.5
HS-21200.334.3345.5−14.9 ± 0.8
HF-1121031.336−15.0 ± 1.1
Intermediate plants
HF-9401.6611.333−14.0 ± 0.7
HS-1970010.664−14.1 ± 0.6
Sulino 400.3310.662.33−14.7 ± 0.6
Winter susceptible plants
HS-43620.3300.660−11.2 ± 0.5
HS-529101.33−11.4 ± 0.5
LSD (P = 0.05)± coefficient interval for P = 0.05

Cold acclimation was also generally accompanied by a decrease in current quantum yield (φPSII) and photochemical quenching (qP) and an increase in nonphotochemical quenching (NPQ). However, in all of these parameters the relations with winter hardiness were less clear. There was a difference in the response of HS-262; it was the only genotype of those studied, including the parent material, in which φPSII and qP increased and NPQ decreased in response to cold hardening. HS-262 was also relatively hardy, with a high spring regrowth score in 2001. There were no differences in accumulation of anthocyanins between winter hardy and winter susceptible plants.

Responses of Fv : Fm, qP and NPQ to high light at low temperature in hardened and unhardened hardy and susceptible plants

Fv : Fm declined in the winter susceptible plants with time of exposure to high light with only one transient difference between hardened and unhardened material (Fig. 4). Thus photoinhibition increased with time and there was little acclimation during cold hardening. In the winter hardy plants Fv : Fm declined for only the first 2 h, with hardening reducing the initial Fv : Fm and its rate of change, giving higher final values than in unhardened material. Thus the winter hardy plants showed less photoinhibition, and cold hardening increased their ability to avoid photoinhibition.

Figure 4.

Time course of changes in chlorophyll fluorescence from first exposure to 1200 µmol m−2 s−1 PPFD at +2°C in two winter hardy (HF-105, open and closed circles; HS-262, open and closed triangles) and two winter susceptible (HS-436, open and closed squares; HS-529, open and closed diamonds) genotypes either nonacclimated (open circles, open triangles, open squares, open diamonds, 20/17°C, 200 µmol m−2 s−1) or cold-acclimated (closed circles, closed triangles, closed squares, closed diamonds, +2°C, 300 µmol m−2 s−1) before exposure. Mean of 10 replications ± SE.

In winter susceptible genotypes qp rose in the first 2 hr of high light treatment, irrespective of preceding cold acclimation (Fig. 4). The same was observed in nonacclimated winter-hardy genotypes. In the susceptible material the subsequent changes were quite variable, with contrasting directions of response in the two genotypes by the end of measurements. In cold acclimated winter hardy material qp did not change during high-light treatment and in nonacclimated plants further changes in qP were in opposite directions in the two genotypes. It can be supposed that cold acclimation of winter hardy plants makes photochemical quenching more stable during high light treatment.

The changes in NPQ on exposure to high light were much more rapid and variable in the susceptible than the hardy genotypes. In the susceptible genotypes there was a greater overall fall in NPQ and cold hardening resulted in a higher NPQ by the end of measurements. In the hardy material NPQ fell more slowly and was eventually higher than in the susceptible plants.


In these Festuca–Lolium cultivars and their androgenesis-derived progeny, variation in winter survival and freezing tolerance are connected, at least partially, with variation in resistance of plants to cold-induced photoinactivation of PSII. Under field conditions, winter hardy plants were characterised by lower maximum quantum yield of PSII (Fv : Fm), with no differences in photochemical efficiency (φPSII) in the autumn. The correlation between winter survival and Fv : Fm was very high (−0.73 and −0.67 in two successive years), particularly taking into account the large number of traits associated with winter hardiness. However, the range of Fv : Fm in these experiments is relatively low and typical of those often quoted for healthy leaves. Thus these changes could be caused by changes in light absorbance or in the spill-over from PSII to PSI rather than photosynthetic acclimation. However, if either of these were large factors the changes in NPQ and qp in the more extreme genotypes studied in detail would be much smaller. The most common protective change here was increased nonphotochemical quenching. Various mechanisms have been proposed for this (Huner et al., 1998), but the most likely would be increased energy dissipation of excess light energy via the xanthophyll cycle (Demmig-Adams & Adams, 1992; Huner et al., 1993).

During cold acclimation, these androgenic plants showed variability in the acclimation of photosynthetic apparatus to low temperature. As well as the decrease in Fv : Fm, there was an increase in nonphotochemical quenching at fixed irradiance in all genotypes except HS-262, which increased photochemical quenching, an alternative path for managing excess energy. Both winter hardy plants studied in the experiment showing the effects of high-light treatment became more tolerant to PSII photoinactivation than winter-susceptible plants after cold acclimation, but the mechanisms involved differed. HS-262 showed again the increased light energy consumption in photochemical processes. HS-262 also accumulated anthocyanins, although they may not be very effective in protecting these Festulolium genotypes from photoinhibition as they were also found in the winter susceptible genotypes and were not present in three of the winter hardy genotypes. The lack of an obligate link between anthocyanin and cold hardening in herbaceous plants is consistent with the results of Leyva et al. (1995), who demonstrated that Arabidopsis mutants unable to accumulate anthocyanins were able to cold acclimate to levels similar to the wild-type. By contrast, Havaux & Kloppstech (2001) demonstrated that flavonoid-deficient mutants were high-light sensitive but this was not related to freezing tolerance in their report. The presence of more than one mechanism of photoinhibition avoidance means that if only simple screening for one protective parameter is used, such as Fv : Fm, then important material that has alternative mechanisms of photoprotection may be discarded. For example selecting for cold tolerance on the basis of a low Fv : Fm would have rejected HS-262 with its high Fv : Fm, although it has high cold tolerance related to protection from photoinhibition by high photochemical quenching. This underlines the importance of measuring several parameters, ideally in unhardened and hardened material as has been done here, to determine which mechanisms of avoidance of photoinhibition are operating. Selection within each mechanism for the highest value of the relevant parameters can then be carried out.

These results emphasise the important role of the photoprotection of the photosynthetic apparatus during cold-acclimation in determining winter hardiness of plants. Although frost resistance and resistance to photoinhibition seem to be independent components of winter hardiness (along with various others, such as tolerance to snow cover), in the germplasm studied in these experiments both frost resistance and the resistance to photoinhibition contribute strongly and almost equally to the variation in winter hardiness. In fact the protection of photosynthetic apparatus in low temperatures is important for proper development of freezing tolerance during cold acclimation and both cold acclimation and the acclimation of photosynthetic apparatus to high-light are partially triggered by the same signal perception and transduction pathways connected with changing redox state of PSII (Huner et al., 1998; Ndong et al., 2001). Previous studies conducted in other Poaceae (wheat, rye) and Brassicaceae showed that cold acclimation in these herbaceous plants was accompanied by increased photosynthetic capacity (Huner et al., 1993, 1998; Rapacz, 1998a,b; Pocock et al., 2001; Savitch et al., 2002). Furthermore, the accumulation of reserves during hardening frequently correlates with increased cold tolerance. Successful recovery from freezing stress depends on the plant's ability to produce new leaves from meristems that have survived freezing temperatures. In Lolium recovery is associated with growth from previously quiescent subapical meristems, because the main apical bud invariably dies at temperatures below −5°C (Harrison et al., 1997). Here, in the majority of these Festulolium lines, in contrast to previous reports in the Poaceae the acclimation of photosynthesis was via increased nonphotochemical quenching. This implies that in these genotypes, some of which were at least as winter hardy as the line that had increased photochemical quenching, the ability to protect the photosynthetic apparatus on overwintering leaves was more important than maximising photochemical quenching in the cold. Thus, further studies of the sources of carbon for the regrowth from the quiescent subapical meristems in these lines with contrasting mechanisms of acclimation would be extremely interesting. However, it should not be forgotten that other mechanisms of protection of the photosynthetic apparatus that will not be detected by chlorophyll fluorescence measurements can occur. For example, following 4 d cold acclimation, increased expression of psbA, which encodes for the D1 protein of PSII, was observed in Festuca pratensis (Canter et al., 2000). This further indicates the importance of repairing damaged PSII reaction centres and their contribution to improved winter hardiness in Lolium and Festuca.

The present work confirms the value of androgenesis for complex trait analysis. Despite being derived from pollen mother cells, the chloroplasts of the androgenic plants were able to function and to acclimate effectively to the various winter stresses. From a small number of parents, populations were generated from independent segregation of intact or recombined chromosomes in a way that tends to increase expression of a range of alleles relevant to winter hardiness traits. Each androgenic plant has a unique combination of genes, which is unlikely to be recovered in another androgenic plant. Thus, here we have effectively ‘dissected’ the components of the parent plant's genome that are responsible for the maintenance of an effective photosynthetic apparatus under winter stresses. It is possible that the parent species, F. pratensis and L. multiflorum have genes governing different mechanisms of photoinhibition, and that these had been separated between plants within the androgenic populations derived from the Festulolium cultivars. Currently, it is not known whether this ‘dissection’ is the result of the removal of competing mechanisms or just a change in regulation, analogous to the change in acclimatory mechanisms between sun and shade adapted leaves in Vinca minor (Adams et al., 2002). It revealed mechanisms of protection from photoinhibition (increased photochemical quenching) that were not expressed in the parental cultivars, but that were known in other members of the same family, Poaceae. This confirms the ability previously reported (Lesniewska et al., 2001) of androgenesis to reveal traits present but not expressed in the parent material.


This study was performed with the financial support of European Commission: SAGES project (QLK5–CT−2000–00764). We thank M. S. Dhanoha for advice on the use of bootstrap techniques for determining confidence limits of ratios.