CO2 enrichment predisposes foliage of a eucalypt to freezing injury and reduces spring growth



    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author
    • *

      Present address: Department of Biological Science, University of Nevada, Las Vegas NV 89154–4004, USA.


    Corresponding author
    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author

    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author

    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author
    • Present address: St Mary's College of Maryland, Department of Biology, Schaefer Hall 18952 E. Fisher Road, St. Mary's City, MD 20686–3001, USA


    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author
    • Present address: St Mary's College of Maryland, Department of Biology, Schaefer Hall 18952 E. Fisher Road, St. Mary's City, MD 20686–3001, USA


    1. CRC for Greenhouse Accounting and Ecosystem Dynamics Group, Research School of Biological Science, Australian National University, GPO Box 475, Canberra ACT 2601, Australia
    Search for more papers by this author

Beth R. Loveys. Fax: +61 (2) 6125 5095; e-mail:


Seedlings of Eucalyptus pauciflora, were grown in open-top chambers fumigated with ambient and elevated [CO2], and were divided into two populations using 10% light transmittance screens. The aim was to separate the effects of timing of light interception, temperature and [CO2] on plant growth. The orientation of the screens exposed plants to a similar total irradiance, but incident during either cold mornings (east-facing) or warm afternoons (west-facing). Following the first autumn freezing event elevated CO2-grown plants had 10 times more necrotic leaf area than ambient CO2 plants. West-facing plants had significantly greater (25% more) leaf damage and lower photochemical efficiency (Fv/Fm) in comparison with east-facing plants. Following a late spring freezing event east-facing elevated CO2 plants suffered a greater sustained loss in Fv/Fm than west-facing elevated CO2- and ambient CO2-grown plants. Stomatal conductance was lower under elevated CO2 than ambient CO2 except during late spring, with the highest leaf temperatures occurring in west-facing plants under elevated CO2. These higher leaf temperatures apparently interfered with cold acclimation thereby enhancing frost damage and reducing the ability to take advantage of optimal growing conditions under elevated CO2.


It is well established that plants grown under elevated CO2 exhibit enhanced photosynthesis and biomass production at least in the short to medium term (see Drake, Gonzalez-Meler & Long 1997). However the impact of the complex interaction between timing of light interception, temperature and elevated CO2 on plant photosynthesis and growth is less well documented. With the predicted changes in climate, specifically, doubling [CO2] and increasing temperature combined with more frequent temperature extremes that are likely to occur over the next century (Houghton et al. 2001) it is important to better understand how these environmental parameters will affect plant growth, survival and distribution. To this end we aimed to address the question: does exposure to elevated CO2 alter physiological and growth responses to light and temperature?

Plant responses to elevated CO2 are altered by temperature (Long 1991). The temperature dependence of both the rubisco-limited and electron transport-limited components of photosynthetic assimilation rates (A) are well known (Farquhar, von Caemmerer & Berry 1980; Medlyn et al. 2002). At lower temperatures A is limited by rubisco activity whereas at higher temperatures the rate of electron transport is limiting, although other components may also be affected at temperatures above 30 °C (e.g. increasing RuBP pool and decreasing activation of rubisco) (von Caemmerer 2000). The temperature response curve of A also changes with light intensity; over a 10-fold range in PPFD there is very little difference in A at temperatures under 10 °C, whereas A increases significantly with increasing PPFD at higher temperatures, up to 30 °C (von Caemmerer 2000). These components will also be differentially affected by [CO2]. As [CO2] rises carboxylation rates increase as CO2 competitively inhibits the oxygenase reaction of rubisco. Therefore, photorespiratory CO2 production will be reduced, thus increasing net CO2 uptake (Farquhar et al. 1980). Long (1991) showed that light-saturated assimilation rates will be much more responsive to [CO2] at warmer leaf temperatures (35% increase under elevated CO2 at 30 °C) than at low leaf temperatures (4% increases under elevated CO2 at 10 °C).

Previous studies have shown that plants may suffer the detrimental effects of cold and heat stress to a greater extent when grown under elevated CO2 (Roden & Ball 1996; Lutze et al. 1998; Beerling et al. 2002). During warm periods that are optimal for loblolly pine growth, the efficiency of electron transport through photosystem II (PSII) was higher in elevated CO2 than ambient CO2-grown plants, but during the winter the situation was reversed as elevated CO2 plants had lower electron transport rates (Hymus et al. 1999). Elevated CO2 may amplify the interactive effects of low temperature and high light accelerating the rate of de-acclimation in freeze tolerance during spring thereby leaving plants more susceptible to frost damage events (Lutze et al. 1998; Terry, Quick & Beerling 2000). Studies of electrolyte leakage and ice-nucleation on evergreen and deciduous species have demonstrated significant increases in the minimum temperature at which leaf damage occurs when plants grown under elevated CO2 are subjected  to  freezing  temperatures  (Lutze  et al.  1998; Terry et al. 2000; Beerling et al. 2001). Evaluation of plant performance under elevated CO2 must take into account temperature-dependent effects.

To separate the interactive effects of elevated CO2 at high light and variable temperature conditions throughout the day and season, seedlings of the broad-leaf evergreen Eucalyptus pauciflora Sieb. Ex Spreng (snow gum) were grown within open-top chambers. The chambers were fumigated with air containing ambient and 2× ambient [CO2], and seedlings were divided into east- and west-facing populations using shade screens. It was hypothesized that elevated CO2 would amplify the effects, beneficial or detrimental, of east- or west-facing plants. It was predicted that: (1) east-facing plants would be less able to take advantage of the high light in the early morning due to the low temperatures, and would be at greater risk of frost damage; and (2) west-facing plants would be more likely to benefit from high light when temperatures are higher and would be protected from frost damage to a greater extent than east-facing populations. A third hypothesis was developed following a serendipitous finding that an early autumn freezing event resulted in significantly greater leaf damage under elevated than ambient CO2 particularly in west-facing plants. Therefore it was hypothesized that: (3) growth under elevated CO2 resulted in less cold acclimation and greater susceptibility to freezing damage because of higher leaf temperatures due to reduced stomatal conductance.


Plant material and growth conditions

Seeds of Eucalyptus pauciflora Sieb. Ex Spreng. were collected from trees in the Gudgenby Valley, Australian Capital Territory (ACT), elev. 1000 m. Seeds were cold-stratified under moist conditions at 4 °C for 4 weeks before germinating on sand in a glasshouse. Seedling were grown for 2 weeks and then selected for uniformity of size and transferred to seedling tubes and grown outside under natural conditions in Canberra (elev. 600 m) for 3 months. After this time the plants were again selected for uniformity of size and leaf number, and transferred into larger (15 cm × 40 cm) open-bottom pots which were buried in the ground within open-top chambers (OTCs) and the soil surface around the pots was covered with a layer of straw. The pots were filled with soil from the surrounding area. The experiment began on 22 March 2001 and the final harvest was performed on 27 November 2001. However, a subsample of plants was left in the chambers until January 2002 to continue non-destructive measurements.

Open-top chambers

In a pasture near Bungendore in south-eastern Australia (35°15′ S, 149°27′ E; elev. 700 m), 10 OTCs were installed as five replicate pairs, flushed with air containing either ambient or 2× ambient [CO2]. The [CO2] in the chambers was monitored periodically using a portable gas exchange system (LI 6400; Li-Cor Inc., Lincoln, NE, USA) and averaged 652 µmol mol−1 (95% confidence interval, 25.5 µmol mol−1). The flow though the chambers was 1.61 m3 min−1. Chambers were as described by Ashenden, Baxter & Rafarel (1992) and modified by Lutze et al. (1998) and Roden, Egerton & Ball (1999). Chambers were further modified for the current experiment by the placement of 10% transmittance shade screens in the middle of the chambers at a height of 0.5 m. These screens separated the plants (5–7 per side) into east- and west-facing populations that received direct light in either the morning or afternoon, respectively. The leaves of all plants were constrained using small wooden skewers to face either the east, for east facing plants or west, for west facing plants. This ensured that there was as little shading as possible.

Assessment of leaf frost damage

Necrotic leaf area was assessed using visual criteria. The leaf area of every leaf was determined using allometric data and the percentage of the necrotic area was estimated. Frost damage was assessed on 25 April following the first significant frost on 13 April and again on 26 September to determine how the freezing damage had effected the plants during the winter.

Chlorophyll fluorescence

The chlorophyll a fluorescence parameter Fv/Fm was measured weekly on exposed leaves from each side of the divided chambers, using a portable plant efficiency analyser (PEA; Hansatech, King's Lynn, UK). Following 30 min dark adaptation five replicate plants in each orientation and gas treatment were measured at 1030 h and then again at 1530 h.

Thermocouple measurements of leaf and air temperature

Temperature was measured on one leaf and of the adjacent air (less that 5 cm from the measurement leaf) in each treatment combination in three of the chamber pairs. Temperatures were measured with copper–constantan thermocouples (64 µm diameter) referenced against a PT-100 platinum resistance thermometer. Thermocouples were attached to the leaf lamina on the inward facing surface using a small piece of surgical tape, so were not exposed to any direct sunlight. Thermocouples were scanned every 10 s and a 15-min average was recorded on a DT 100 data-logger (Data Electronics, Rowville, Victoria, Australia). Reference measurements were made in a weather screen 1.2 m above the ground located on the western side of the study site. Precipitation was recorded from weekly readings of a rain gauge near the weather screen.

Gas exchange measurements

Photosynthetic CO2 assimilation rates at saturating irradiance (Asat) 1000 µmol m−2 s−1 were measured in the field with a portable gas exchange system (LI 6400; Li-Cor Inc.), fitted with a red/blue LED light source (6400–02B). Light saturation had been previously determined by Egerton et al. (2000). Measurements were made on one leaf from east- and west-facing plants in every chamber pair on 26 June, 11 September and 22 November 2001. East-facing leaves were measured between 0830 and 1100 h, whereas west-facing leaves were measured between 1300 and 1530 h. Times were chosen such that the integrated light dose experienced for east- and west-facing plants was near equal. The same leaves were used for CO2 assimilation measurements as had been used to measure chlorophyll a fluorescence. To estimate plant carbon gain measurements were made at ambient temperature and at growth [CO2] (either 350 or 700 µmol mol−1) in the cuvette.

Leaf spectral properties

Leaf transmittance (T) and reflectance (R) was measured and absorptance calculated (A: A = 100 − %T − %R) every 5 nm between 350 and 1100 nm using a portable spectroradiometer with an external integrating sphere connected by a quartz fibre-optic cable (LI-1800 with LI-1800–12S; Li-Cor Inc.). In spring (late October and November), measurements were made on young leaves and also on mature leaves from the previous year's growth under the two CO2 treatments. In autumn (February to April) young plants that had developed outside the chambers at ambient CO2 were transplanted into the chambers, and leaf spectra measured on mature leaves subsequently developed within the chambers 6 weeks later.

Stomatal density measurements

From plants growing in treatment for 11 months one youngest fully expanded leaf from each treatment combination in each block was selected in January 2002. Clear acrylic varnish was painted in a thin layer on both abaxial and adaxial surfaces of the attached leaves, creating a mirror image of the leaf surface. Once dry the layer of varnish was removed using clear adhesive tape, both varnish and tape were placed on a glass microscope slide and viewed under a binocular microscope at 40× objective magnification. Stomata were easily distinguishable from oil glands. Stomatal numbers in three distinct regions on each leaf surface were counted (n = 15).


The randomised split plot design was utilized for the purpose of statistical analysis. Data were analysed by two-way analysis of variance (Anova) and appropriate post-hoc tests using Genstat (v. 6) (VSN, Hemel Hempstead, UK).


Environmental conditions

The daily minimum and maximum temperatures at the field site were recorded throughout the duration of the experiment, March to November 2001 (i.e. beginning of autumn to the end of spring) (Fig. 1a). On 13 April the first freezing event of the year occurred with a low of −5.8 °C, temperatures rose to an afternoon high of 20.4 °C giving a diurnal temperature range for that day of 26.2 °C. The average daily maximum temperature was 16 °C and the average minimum temperature was 2.1 °C, with the last freezing event of the year taking place on 22 October During the experimental period 330 mm of rain fell, slightly below the long-term average (Fig. 1b). East- and west-facing leaves intercepted similar integrated daily irradiance (Fig. 1c).

Figure 1.

(a) Daily screen air temperature and (b) weekly rainfall from February to December 2001 at Bungendore, NSW and (c) an example of a daily time course of photosynthetic photon flux density (PPFD) measured on the east- (dotted line) and west-sides (solid line) of one OTC on 29 October 2001. The integrated light received on that day on the east-side was 29.28 mol m−2 d−1 and on the west-side was 30.59 mol m−2 d−1.

Photosynthetic efficiency and canopy loss following early season frost

Photosynthetic efficiency, Fv/Fm was depressed following the 13 April frost (Fig. 2a). Fv/Fm was 27% lower (P = 0.05) for west-facing elevated CO2 plants than the average of the other treatments when measured 4 d after the frost event. There was no difference in Fv/Fm between east-facing ambient CO2 and elevated CO2 plants.

Figure 2.

(a) Mean chlorophyll a fluorescence (dark-adapted Fv/Fm) measured in the morning for east-facing and in the afternoon for west-facing plants; (b) median leaf death as a percentage of total canopy area; and (c) median living leaf area of plants following the 13 April frost of east- and west-facing plants grown under elevated CO2 and ambient CO2. n = 5, bar = SEM.

Leaf damage was estimated 12 d after the frost event that caused foliar necrosis (Fig. 2b). Median leaf damage was significantly higher (P < 0.001) for plants grown under elevated CO2 in comparison with ambient CO2-grown plants (52 and 4% canopy loss, respectively). West-facing plants experienced, significantly greater damage (67% leaf damage) in comparison with east-facing plants (37% leaf damaged, P = 0.03). The same pattern was revealed in the ambient CO2-grown plants yet the loss of leaf tissue was considerably lower (5% loss for west-facing and 3% loss for east-facing plants).

At the end of winter, surviving leaf tissue was estimated non-destructively to compare the net effect of frost damage on elevated CO2, vis-à-vis ambient CO2-grown plants (Fig. 2c). Elevated CO2-grown plants had significantly less surviving leaf area than plants grown under ambient CO2 (P = 0.02). Further, the west-facing elevated CO2 plants had significantly less leaf area than the east-facing plants (P = 0.03).

Leaf temperature

The surprising result described above, namely that frost damage was greater in west- than east-facing plants at elevated CO2, implies that difference in leaf temperature could affect acclimation to freezing. Thermocouples were placed on leaves to determine if plants growing under elevated CO2 were warmer than ambient CO2-grown plants. Figure 3a and b show leaf temperature for a single leaf in each of the treatments over the course of a single day (5 July 2001). East-facing plants had higher leaf temperature under elevated than ambient CO2 during exposure to direct morning sunlight. Similarly, west-facing plants had higher leaf temperature under elevated than ambient CO2 during exposure to direct afternoon sunlight.

Figure 3.

Diurnal temperature measurements of a single leaf growing under elevated CO2 (solid circles) or ambient CO2 (clear circles) of (a) east-facing or (b) west-facing plants. Data are from 5 July 2001.

The difference in leaf minus air temperature (ΔT) (between 1100 and 1300 h on the east and 1300 and 1500 h on the west) was calculated and the observation divided into temperature classes. This made it possible to establish the most commonly observed difference in maximum leaf and air temperature (ΔTm) (Fig. 4a & b) and to compare the distribution of differences between leaf and air temperature for all measurements made during each day (ΔTa) for each of the treatments (Fig. 4c & d). ΔTm was calculated from the absolute maximum leaf temperature minus the corresponding air temperature at that time. For east-facing plants the majority of observations of ΔTm for both elevated CO2 and ambient CO2 treatments fell into the 1 °C class (Fig. 4a). For west-facing plants the greatest number of observations for the ambient CO2-grown plants fell into the 2 °C temperature class whereas for elevated CO2-grown plants most of the observation were in the 3 °C class (Fig. 4b). When all data (ΔTa) were considered, leaf temperature was most frequently 1 °C above ambient air temperature (Fig. 4c & d). However, there was more frequent occurrence of leaf temperatures ranging from 2 to 5 °C above ambient air temperature under elevated CO2, with the effect being most pronounced in west-facing plants.

Figure 4.

Observations of maximum leaf minus corresponding air temperature at a given time (ΔTm) for all chambers (n = 5) were classified into temperature classes for elevated CO2 (solid bars) and ambient CO2 (clear bars)-grown plants, east-facing (a) and west-facing (b). The difference between all corresponding leaf and air temperature measurements (ΔTa) was also analysed this way for elevated and ambient plants east-facing (c) and west-facing (d). As with Fig. 3 east data includes 1100–1300 h records while west data includes 1300–1500 h.

Stomatal density and leaf spectral characteristics

As leaf temperature was affected by our treatments we examined two sets of leaf attributes that may influence leaf temperature: stomatal density and leaf spectral characteristics. It has been observed that stomatal density can change when plants are grown in elevated CO2. Initial observations suggested that stomatal density declined in a CO2-enriched atmosphere (Woodward 1987); however, accumulated observations show that plant responses are likely to be species specific. If stomatal densities were different between our treatments this may affect leaf temperature. In our study stomatal densities were unchanged by either growth [CO2] or aspect: west elevated CO2–, 218.9 µm−2 ± 22.3, west ambient CO2–, 237.5 µm−2 ± 25.5, east elevated CO2–, 235.3 µm−2 ± 21.0, east ambient CO2–, 228.1 µm−2 ± 11.7.

There were no differences in the optical properties of leaves that might contribute to temperature variation between treatments. The only spectral difference between leaves developed under the two CO2 treatments was a very small and transient decrease in transmittance at 555 nm for young leaves on the elevated CO2 plants; this difference was evident in spring (October) but disappeared when the same leaves were re-measured 2 weeks later. By that stage, spectra taken for leaves developed under the two regimes were identical, as they were for mature leaves developed the previous year. For both young and mature leaves, and for both CO2 treatments, the abaxial surface reflected slightly more radiation than the adaxial surface; this difference was most pronounced in the visible portion of the spectrum, but never more than a few percent.

Stem elongation and bud break

West-facing plants under elevated CO2 were the slowest to respond with stem elongation and bud-break to increasing temperature in spring; this was probably a consequence of the severe frost damage suffered at the beginning of winter (Fig. 5a & b). Bud break of west-facing elevated CO2 plants lagged behind ambient CO2-grown plants throughout the measured period. However, by mid-November west-facing plants were able to take advantage of the benefits of growing under elevated CO2 and stem elongation matched that of the ambient CO2-grown plants. Stem elongation and bud break (Fig. 5a & b) of east-facing plants under elevated CO2 was adversely affected by a late spring frost on 3 October (Fig. 1a) after which time plant growth of east-side plants greatly decreased.

Figure 5.

Number of buds (a) and total stem length (b) of plants growing under all treatments; elevated CO2 east-facing (solid triangles), ambient CO2 east-facing (clear triangles), elevated CO2 west-facing (solid circles) and ambient CO2 west-facing (open circles) n = 5, bar = SEM.

Seasonal changes in photochemical efficiency

The values of Fv/Fm decreased with lower temperature and the approach of winter, and increased with rising temperature during the spring (c.f. Figs 6 & 1a). However, during autumn Fv/Fm was consistently lower for elevated CO2 than ambient CO2-grown plants of either orientation. As temperatures increased from winter into spring the Fv/Fm of east-facing plants under elevated CO2 failed to recover following a late frost in September (Figs 1a, 6a & c). In contrast the Fv/Fm of west-facing elevated CO2-grown plants recovered and by mid-September was not different than ambient CO2-grown plants (Fig. 6b & d).

Figure 6.

Seasonal patterns of intrinsic efficiency of photosystem II (dark-adapted Fv/Fm) for elevated CO2 (solid circles) and ambient CO2-(clear circles) grown plants. For the first 2 weeks of measurement at the end of April Fv/Fm was only recorded on west-facing plants in the afternoon and on east-facing plants in the morning; by May measurements were taken in the morning and afternoon for both the east-facing (a and c) and the west-facing plants (b and d). n = 5, bar = SEM.

Carbon assimilation and stomatal conductance

When measured at growth [CO2], light-saturated rates of assimilation (Asat) for the east-facing plants were not affected by [CO2] suggesting that down-regulation of photosynthesis under elevated CO2 had occurred (Fig. 7). The Asat of west-facing plants grown at elevated CO2 was higher than the Asat of plants growing at ambient CO2 for all measurement periods. Stomatal conductance (gs) was generally lower in plants grown at elevated CO2 in comparison with ambient CO2 plants. For the most part west-facing plants had lower gs than east-facing plants. Stomatal conductance was significantly lower in winter and early spring (P < 0.05) for elevated CO2-grown plants but not in late spring when gs was considerably higher, probably due to increased water availability and temperature (Fig. 7).

Figure 7.

Light-saturated photosynthesis (Asat) and stomatal conductance (gs) measured in winter (a and b), early spring (c and d) and late spring (e and f) for plants growing under elevated CO2 (solid bars) and ambient CO2 (clear bars) east-facing and west-facing. n = 5, bar = SEM.

Biomass allocation and growth

By the final harvest at the end of spring (November) biomass accumulation indicated that the west-facing plants growing under elevated CO2 had not completely recovered from the damage suffered from the initial freezing event in April (Fig. 8). Although leaf area was somewhat lower for elevated CO2 compared to ambient CO2-grown plants, root dry mass was significantly lower (P = 0.03), there was also a moderately significant CO2 by aspect interaction (P = 0.06). The significant reduction in root biomass accounted for half of the reduction in total biomass that was apparent in the west-facing elevated CO2 plants. For east-facing plants there was no difference in biomass accumulation by late spring between elevated CO2 and ambient CO2-grown plants.

Figure 8.

Final harvest in late spring (November 2001), leaf area (a), shoot dry weight (b), root dry weight (c) and total plant dry weight (d) for elevated CO2 (solid bars) and ambient CO2 (clear bars) east- and west-facing plants. n = 5, bar = SEM.


This study shows that the enhanced growth that is often observed under elevated CO2 in controlled environment studies may not always be exhibited when plants are exposed to other environmental factors that can constrain growth. The interaction between changes in physiology that occur as a result of exposure to elevated CO2 and other environmental parameters is likely to be complex. Our first two hypotheses that east-facing populations would be less able to take advantage of high light in the early morning due to low temperatures and would also be at greater risk of frost damage than protected west-facing populations was supported, at least during the spring. Elevated CO2 reduced the ability of leaves to resist frost-damage particularly for west-facing plants in autumn (Fig. 2) and resulted in a decline in Fv/Fm for the east-facing plants in spring (Fig. 6). Frost damage lowered photochemical efficiency and reduced leaf area thereby diminishing the ability to take advantage of optimal growing conditions during spring, ultimately decreasing biomass production. The greater rate of carbon assimilation throughout the year of west-facing plants under elevated CO2 (Fig. 7) did not compensate for tissue loss and the subsequent constraint on growth rate.

The increased freezing damage of west-facing elevated CO2 leaves was unexpected. Our original prediction, based on the well-documented response of the biochemistry of photosynthesis to temperature, light and [CO2] was that west-facing leaves would be less likely to suffer freezing damage. Protected from the cold, high light conditions that are experienced by east-facing plants, west-facing plants would be expected to take greater advantage of elevated CO2 during the warmer afternoons. Although CO2 assimilation was enhanced for west-facing plants under elevated CO2, they were less prepared for a frost event in autumn and were therefore more susceptible to frost damage. We have previously shown that late spring frost damage occurs due to an acceleration in de-acclimation when snow gum was grown at elevated CO2 (Lutze et al. 1998), and here we show that an apparent lag in acclimation as winter approached may have led to significant foliar damage during autumn frosts.

Our third hypothesis that growth under elevated CO2 reduced stomatal conductance and thus increased leaf temperature, particularly of west-facing plants, leading to changes in timing of acclimation was also supported. Our data suggest that the reduction in gs commonly observed in plants grown at elevated CO2 (Drake et al. 1997; Curtis & Wang 1998; Nowak et al. 2000) was likely to result in higher leaf temperature (Figs 3 & 4) (Siebke et al. 2002) possibly delaying the necessary signal transduction required to cold acclimate evergreen leaves. Acclimation to higher leaf temperature comes at the cost of low temperature acclimation and increased susceptibility to low temperature membrane damage. Data presented here revealed that reduced conductance and greater leaf temperature in west-facing, elevated CO2-grown plants, may have increased their susceptibility to frost damage during autumn, suggesting that the signal for cold acclimation may have been delayed or altered. However, during the spring, it was predicted that warmer leaves may falsely signal the rapid advance of summer to elevated CO2-grown plants, increasing the rate of premature de-acclimation and increasing the chance of frost-induced damage (Lutze et al. 1998). Similarly, seasonal Fv/Fm data (Fig. 6), bud count and stem elongation data (Fig. 5) suggested that east-facing plants growing under elevated CO2 were more susceptible to early spring frosts due to premature de-acclimation.

Spectral and stomatal density data revealed no changes with respect to either growth [CO2] or aspect. It is therefore likely that changes in leaf temperature were due primarily to changes in gs. Phenology is considered one of the most sensitive plant traits to climate change (Sparks & Menzel 2002) and must be considered when assessing the impact of our treatments on freeze tolerance and growth. However leaf turnover for Eucalyptus species is slow and occurs every 1–3 years, with the average leaf age of 18 months for evergreen eucalypts (Chapin 1980; Boland et al. 1984). As our experiment lasted only 8 months it is unlikely that there was any significant leaf turnover or change to leaf phenology that would have influenced our results.

In a natural environment there are additional factors relating to temporal separation of resources that must be taken into account when predicting plant growth. Wayne & Bazzaz (1993) proposed a hypothesis, the resource congruency hypothesis which states that plants often experience high light levels when other resources are limiting, particularly in forest gaps, therefore reducing potential photosynthetic carbon gain. This has important implications for the success of tree seedlings growing on east- or west-facing sides of forest fragments and is likely to result in an asymmetric growth response. The inevitability of the solar track of the sun across the sky results in cold, bright conditions prevailing for east-facing plants that are unlikely to be as favourable for photosynthesis and growth as the warmer high-light conditions characteristic of west-facing populations. In an ecological framework it therefore appears that the process of delayed acclimation is likely to have a greater impact for west-facing plants of forest fragments whereas premature de-acclimation is more likely to affect the east-facing plants in forest fragments.

The literature documenting the interaction between elevated CO2 and exposure to freezing temperatures shows contrasting results. Tolerance to freezing temperatures has been shown to increase in Betula alleghaniensis (Wayne, Reekie & Bazzaz 1998) and Yucca brevifolia, Y. scidigera and Y. whipplei (Loik et al. 2000). Conversely, freezing tolerance declines in Pinus sylvestris (Repo, Hanninen & Kellomaki 1996) Ginkgo biloba (Terry et al. 2000), Eucalyptus pauciflora (Lutze et al. 1998), Vaccinium myrtillus (Taulavuori et al. 1997), Picea mariana (Margolis & Vezina 1990), Bromus erectus, Cirsium acaule, Sangisorba minor, Salvia pratensis and Trifolium medium ( Obrist, Arnone & Korner 2001). Finally there are also several reports of elevated CO2 having no effect on freeze tolerance in evergreen species (Wiemken, Kossatz & Ineichen 1996; Dalen, Johnsen & Ogner 2001; Naumburg, Loik & Smith 2004). Obviously there are species-specific changes in freeze tolerance at elevated CO2 which may be related to the signalling mechanisms that lead to acclimation and de-acclimation of plants to freezing temperatures. For example, some species rely only on the gradual decline of non-freezing temperatures (e.g. E. pauciflora) to initiate cold acclimation (Tibbits & Reid 1987), whereas other species respond to a combination of minimum temperatures and shortening day length (e.g. P. abies) to begin the acclimation process (Dalen & Johnsen 2004). Differences between studies may also be due to different experimental systems and designs as the list above includes results from field studies, open-top chambers and glasshouse experiments. As the mechanisms for acclimation are likely to be a complex combination of physiological and biochemical processes (Browse & Xin 2001), it is not surprising that contrasting responses to multiple factors are observed. If perception of gradually declining temperature is the most important signal to induce freeze tolerance, then the increased leaf temperature that we have reported for west-facing plants growing at elevated CO2 will certainly delay the signal transduction for cold acclimation. However, for species that utilize a combination of environmental signals to initiate cold acclimation, their signal transduction pathways may be less influenced by increases in atmospheric [CO2].


Effects of climate change on phenology for both evergreen and deciduous species through either early onset of bud break in spring, lengthening of growing season or delay in leaf senescence in autumn (see Walther 2003) have been widely interpreted as positive in relation to net CO2 uptake. Specifically, there is likely to be an increase in the duration of time for plants to harness and utilize incident light, therefore potentially increasing global carbon sinks. However, for over-wintering plants, growth under elevated CO2 and the resulting increase in leaf temperatures may cause a delay in the detection of environmental signals to initiate the cold-acclimation process resulting in plants being more susceptible to the inevitable frost events that precede winter in temperate and boreal forests. Therefore increases in carbon gain caused by changes in the timing of phenological events resulting from climate change are likely to be offset by potentially significant loss of photosynthetic tissue due to the enhanced susceptibility of elevated CO2-grown plants to frost damage. This observation has broad-reaching consequences for forest growth dynamics and revegetation practices for disturbed lands.


The authors would like to thank Mr Tim Hobbs and Mr Paul Sillis for permission to conduct experiments on their property, Mr Wayne Pippen for technical assistance and Drs John Evans, Dan Bruhn, Paul Kriedemann and Michael Robinson for comments on earlier versions of this manuscript. Dr John Evans is also thanked for the loan of the spectroradiometer. Many thanks to Dr Dan Bruhn for providing OTC [CO2] data.