Low temperature effects on photosynthesis and growth of grapevine


Robert T. Furbank. Fax: + 61 26246 5000; e-mail: robert.furbank@csiro.au


Growth and photosynthesis of grapevine (Vitis vinifera L.) planted on two sloping cool climate vineyards were measured during the early growth season. At both vineyards, a small difference in mean minimum air temperature (1–3 °C) between two microsites accumulated over time, producing differences in shoot growth rate. The growth rates of the warmer (upper) microsite were 34–63% higher than the cooler (lower) site. Photosynthesis measurements of both east and west canopy sides revealed that the difference in carbon gain between the warmer and cooler microsites was due to low temperatures restricting the photosynthetic contribution of east-facing leaves. East-facing leaves at the warmer microsite experienced less time at suboptimal temperature while being exposed to high irradiance, contributing to an average 10% greater net carbon gain compared to the east-facing leaves at the cooler microsite. This chilling-induced reduction in photosynthesis was not due to net photo-inhibition. Further analysis revealed that CO2- and light-saturated photosynthesis of grapevines was restricted by stomatal closure from 15 to 25 °C and by a limitation of RuBP regeneration and/or end-product limitation from 5 to 15 °C. Changes in photosynthetic carboxylation efficiency implied that Rubisco activity may also play a regulatory role at all temperatures. This restriction of total photosynthetic carbon gain is proposed to be a major contributor to the temperature dependence of growth rate at both vineyards during the early season growth period.


trans-thylakoid pH gradient


light- and CO2-saturated photosynthetic capacity


light-saturated carbon assimilation rate


atmospheric CO2 partial pressure


intercellular CO2 partial pressure


leaf transpiration rate


intrinsic quantum efficiency of PSII photochemistry


heat degree days


stomatal conductance


apparent thylakoid linear electron transport rate


non-photochemical quenching


pulse amplitude modulation chlorophyll fluorescence system


photosynthetically active radiation


relative stomatal limitation


standard error of the mean


air temperature


leaf temperature


vapour pressure difference between leaf intercellular air spaces and the atmosphere.


Chilling sensitive plants are those that suffer a dramatic and often slowly reversible, i.e. damaging, inhibition of plant processes at low temperatures usually below 15 °C (Berry & Björkman 1980). Photosynthetic and growth processes of chilling-tolerant plant species are also restricted by low temperatures but are able to recover relatively rapidly upon return to warmer conditions (Wolfe 1991). Grapevines, which are often grown in cool climates for premium quality wine production, are frequently exposed to chilling temperatures well below 20 °C (Gladstones 1992). Growth cabinet studies have revealed that grape shoot and root growth and fruiting yields are significantly reduced by large decreases in temperature (Buttrose 1969). More recent studies have claimed that both non-acclimated and acclimated grapevines are sensitive to chilling temperatures but these investigations focused principally on the chilling sensitivity of photosynthesis per se (Chaumont, Morot-Gaudry & Foyer 1995; Flexas et al. 1999). Assuming the fluorescence parameter Fv/Fm is a sensitive indicator of damage to the photosynthetic apparatus, then by the above definition, grapevine photosynthesis would appear to be chilling tolerant since a significant, sustained reduction in Fv/Fm has not been previously observed in field-grown grape leaves subjected to chilling (Chaumont et al. 1997; Flexas et al. 1999; Hendrickson et al. 2003). However, none of these studies evaluated the influence of relatively small perturbations in temperature or large diurnal fluctuations imposed on field-acclimated vine growth. Grace (1988) states that in the absence of other limiting factors a 1 °C increase in temperature can increase plant productivity by approximately 10% and that this correlation is strengthened as the altitudinal and latitudinal temperature limits of a species are approached. Thus, a study of this nature would be especially pertinent where grapevines are grown in cool climates close to their temperature limit such that marginal increases in temperature may accumulate over thermal time to produce large improvements in growth. Recent studies have demonstrated a linear relationship between photosynthetic capacity and plant growth, measured as the accumulation of dry weight over time in potted grapevines. Whole canopy photosynthesis was directly correlated with total dry mass accumulation of potted Chambourcin grapevines (Miller, Howell & Flore 1997) and the study by Ferrini, Mattii & Nicese (1995) showed that the relationship between photosynthetic capacity and total dry mass accumulation was linear and that total shoot length mirrored the accumulation of total dry mass. In that study grapevines treated at below optimum temperature (20 °C) exhibited reduced photosynthetic capacity and thus reduced carbon gain and growth (Ferrini et al. 1995).

Thus, analysis of the response of grapevine leaf photosynthesis to low temperatures may provide an important insight into grapevine growth responses to chilling. Chilling temperatures can limit photosynthesis via stomatal closure, inhibition of thylakoid electron transport and photophosphorylation, inhibition of key enzymes in sucrose and starch biosynthesis, Rubisco inactivation and limitation of sink strength and phloem loading (Allen & Ort 2001). When low temperature events physically and metabolically restrict the demand for carbon, supply exceeds demand and photosynthesis is down regulated to correct this imbalance. This down regulation is effected by reduced Rubisco activity or a reduced rate of RuBP regeneration. The latter is brought about a direct limitation of linear electron transport by chronic photo-inactivation, deactivation of stromal bisphosphatases and/or ‘end product limitation.’ End-product limitation or ‘feedback inhibition’ of RuBP regeneration is a result of limited sucrose/starch/amino acid synthesis leading to an accumulation of phosphorylated intermediates, depletion of the inorganic phosphate pool and an inhibition of ATP synthesis (Leegood & Furbank 1986; Sharkey et al. 1986). Photosynthesis modelling has been used to demonstrate end-product limitation of photosynthesis in grapevines under both drought and warm conditions (Maroco et al. 2002) but has not been investigated under low temperature conditions.

Large cool climate vineyards often consist of genetically identical vines (vegetatively cloned from cuttings of the same source vine) that are planted on slopes to allow overnight cold air drainage (Jackson & Schuster 1997). These grapevines are often planted in rows oriented in a north–south direction to achieve similar exposure of the east- and west-facing vine canopy to full sunlight allowing equal fruit maturation times (Jackson & Schuster 1997). The implications of this on grapevine photosynthesis and growth have not been widely studied. A vineyard of this nature is ideal for studying the biological effects of small differences in temperature between two microsites within the same vineyard without any genotypic variation in response. A north–south orientation would also provide further information on the relative photosynthetic contribution of the east- and west-facing canopy halves that would experience peak irradiance at two different times of the day, the morning and afternoon, respectively.

To our knowledge, this is the first study of the potential impact of transient, chilling-induced reductions of photosynthesis on the growth of field-acclimated grapevines. The aim of this study was to investigate possible interactions between marginal microclimatic variation in minimum temperature and large, early season diurnal variation in temperature range and the consequences of such interactions on growth and photosynthetic carbon gain of cool climate, field-grown vines. Studies of glasshouse-grown vines were carried out to elucidate the possible causal mechanisms underpinning any effect of low temperature on photosynthesis. For this study two cool climate grapevine varieties, Riesling and Pinot Noir, were chosen for analysis.


Field-grown plant material

Photosynthesis and growth of two varieties of grapevine (Vitis vinifera L. cvs. Pinot Noir and Riesling) were studied at two vineyards in New South Wales, Australia. The Riesling variety was studied at a vineyard near Canberra (35°1′ S; 149°2′ E; alt. 650 m) and the Pinot Noir variety was studied at a vineyard near Tumbarumba (35°56′ S; 148°E; alt. 740 m). The spur-pruned grapevines at both sites were planted in rows oriented north–south on 10° slopes such that eastern and western halves of the canopy were exposed to direct sunlight in the morning and afternoon, respectively. The Riesling vineyard was planted on a north-west-facing slope whereas the Pinot Noir vineyard was planted on an east-facing slope. Two subpopulations of vines within each vineyard were chosen at random, and were called the upper and lower microsites with respect to their position on the slope. The upper microsites were 12 and 15 m higher in elevation than the lower microsites for the Riesling and Pinot Noir vineyards, respectively. At both sites, five vines at both the upper and lower sites were allocated to a randomized block design for study. For each vine, a pair of exposed east- and west-facing shoots was selected at equal distances from the trunk of the vine. Preliminary leaf, leaf petiole and soil nutrient analysis at the beginning and the end of the two growing periods showed no significant difference between the upper and lower sites at either vineyard.

Long-term climate data for air temperature, relative humidity, rainfall and rain days were obtained from the Tumbarumba and Canberra airport Bureau of Meteorology weather stations (Australian Bureau of Meteorology 2003). The long-term climatic data for both vineyards shows that mean maximum air temperatures are mild, between 10 and 30 °C throughout the year, and that mean minimum temperatures fall below 10 °C throughout the growing season (October to May). Mean daily 0900 h air temperature is consistently lower than mean daily 1500 h temperature for both regions and during the early growth season (October to December) morning air temperatures were clearly within the chilling range. Air temperature increases rapidly during the morning period. It is not surprising to note that mid-afternoon (1500 h) air temperature is very similar to the daily maximum and that both are above 15 °C throughout the grapevine growth season. Relative humidity is high throughout the average year (40–90%). Both sites have warmer and drier afternoon periods. The mean annual rainfall of 986 mm at the Tumbarumba site occurs predominantly in the winter months and the mean annual rainfall in the Canberra region is lower (632 mm) but more evenly distributed throughout the year. Field-grown grapevines, measured throughout the early growth season in this study, did not experience any apparent water stress at either field site due to the high rainfall at the Tumbarumba vineyard and the supplemental watering at the Canberra vineyard (10–12 L per vine per day as required).

Glasshouse plant material

Grapevines (Vitis vinifera L. cv. Riesling) were grown in 10 L pots in an ambient irradiance (maximum daily PAR = 1600 µmol quanta m−2 s−1), temperature controlled (30/20 °C day/night cycle) glasshouse at the Research School of Biological Sciences, Australian National University, Canberra, Australia. Vines were watered to field capacity daily and given complete, slow-release fertilizer.

Field measurements of growth, irradiance and temperature

Relative growth of vines was measured as total shoot length for the vines growing in the upper and lower microsites of both vineyards. Measurements were regularly undertaken from shortly after budburst in early October 2001 (mid-spring) to mid-January 2002 (mid-summer), encompassing the early part of the growth season. Field measurements were undertaken during the 2000/2001 and 2001/2002 growing seasons for Riesling and Pinot Noir, respectively.

Air and soil temperature at both upper and lower microsites was continuously measured with thermistors (Tiny talk; Gemini Data Loggers, Chichester, West Sussex, UK) and an average recorded every 15 min with a data logger from 1 August 2001 to 17 January 2002 and from 1 August 2000 to 31 January 2001 for the Pinot Noir and Riesling vineyards, respectively. Thermistor probes (6 cm long, 3.3 mm diameter) were placed just above the canopy to measure air temperature and were covered with a Styrofoam cup to prevent heating due to the direct exposure to radiation. Soil thermistor probes (12 cm long, 3.3 mm diameter) were embedded vertically into the soil surface and watered in to ensure full contact. Daily maximum and minimum temperatures were taken from the logged data sets. Cumulative air temperature was calculated according to the following simple equation:


where HDD is heat degree day accumulation in °C day, n is the number of days, maxima are the mean daily maximum air temperatures in °C and minima are the mean daily minimum air temperatures in °C. In this study there was no base temperature for vine growth applied to HDD calculations. A previous study has demonstrated that base temperatures for both budbreak and leaf appearance not only vary widely between grapevine varieties but were also found to fall well below 10 °C (Moncur et al. 1989).

Mean daily minimum and maximum air and soil temperatures (averaged over a month) for the upper and lower microsites for the Pinot Noir vineyard is shown from October to January, which covered the period from budburst to the end of field measurements (Fig. 1). All mean daily air and soil temperatures at the Pinot Noir vineyard site were affected by the time of year (P < 0.001) but independent of microsite by time interaction. In general, air and soil temperature maxima and minima increased as expected as the season progressed from spring to summer at both vineyards (Fig. 1). Mean daily minimum air temperatures at the Pinot Noir vineyard were the exception, showing a decline in January 2002. This may be an artefact of reduced sampling (17 d) in January 2002.

Figure 1.

Mean daily minimum (closed symbols) and maximum (open symbols) air, Tair (a) and soil, Tsoil (b) temperatures for the upper (•, ○) and lower (▪, □) microsites plotted against month of year for the Pinot Noir field site (740 m). Data represents the early growth period from October to January in 2001/2002. Each point is the mean ± SE of 17–31 measurements.

In the absence of strong wind, overnight cold air drainage down the slope resulted in lower overnight temperatures at the lower site of each vineyard. At the Pinot Noir vineyard, the only temperature difference between the upper and lower sites mean daily minimum air temperature (P < 0.001), the upper site being on average 2.9 ± 0.17 °C higher than the lower site (Fig. 1). This trend was consistent for each month measured.

At the Riesling vineyard during the early 2000/2001 season, mean daily minimum air temperature at the upper microsite was 0.8 ± 0.2 °C higher (P < 0.05) than the lower microsite; however, occasionally this difference could be as large as 2.5 °C on any given day (data not shown). This was due to overnight cold air drainage down the slopes of both vineyards. During the day, however, when air was well circulated, there was no difference between microsites for maximum air temperature except during January when the upper microsites were 2.5 ± 0.3 °C higher than the lower sites. Minimum soil temperatures were not different between microsites during the early growth period for the Riesling vineyard (data not shown). Although maximum soil temperatures were 0.58 ± 0.06 °C higher for the upper microsite (P < 0.05) this difference did not accumulate to the same extent as air temperature.

Measurement of photosynthetic parameters in the field and in the laboratory

Infra-red gas analysis

In all laboratory and field measurements of photosynthesis, the youngest, fully expanded leaves from the exposed outer layer of the canopy were used. Gas-exchange parameters were calculated as described by von Caemmerer & Farquhar (1981).

In the field, leaf photosynthetic parameters were measured every two hours during the day using a portable, open circuit, infrared gas analysis system (LI-6400, Li-Cor Inc., Lincoln NE, USA). Simultaneous measurements of CO2 and H2O vapour flux, air (Tair) and leaf (Tleaf) temperature allowed calculation of leaf carbon assimilation (A), stomatal conductance (gs), transpiration (E), intercellular CO2 partial pressure (Ci), and leaf to air vapour pressure difference calculated from Tleaf (vpdL). Artificial illumination was supplied to the leaf from a red-blue LED light source attached to the sensor head. The irradiance was set according to diurnal measurements of incident PAR with a hand-held quantum sensor (LI-190 SA quantum sensor; Li-Cor Inc.). Irradiance was measured by placing the quantum sensor at the leaf surface that was at an angle of 45° from vertical, the average leaf orientation of a split canopy trellising system. Both sides of the canopy were measured throughout the day such that the passage of sunlight from east to west resulted in two distinct morning and afternoon peaks of irradiance for the east- and west-facing halves of the canopy, respectively. An estimate of daily net carbon gain for exposed leaves of the grapevine canopy was determined by integrating a complete diurnal carbon assimilation data set and expressing it on a per leaf area basis. An approximation of total daily net carbon gain was obtained by simply summing the daily carbon gain of east- and west-facing leaves.

In the laboratory, photosynthetic gas exchange measurements were made using an open circuit, temperature-controlled, infrared gas exchange system as described in (Laisk & Oja 1998). Sample and reference gas mixtures of CO2, O2 and H2O vapour (balance N2) were measured using an infrared gas analyzer (LI-6262; Li-Cor Inc.) and an Ametek S-3a O2 analyser (Ametek, Paoli, PA, USA). Leaves were placed in a 43 × 43 × 3 mm chamber, and the upper surface sealed with starch paste to a thermostatted glass window to facilitate heat exchange between the water bath and the leaf. Leaf gas exchange was unperturbed since grapevine leaves are hypostomatous. The leaf chamber was uniformly illuminated from a perpendicular custom-built fibre-optic light guide connected to halogen lamps (Schott KL 1500; Schott UK Ltd, Stafford, UK) and a chlorophyll fluorimeter (PAM 101, H; Walz, Effeltrich, Germany).

Leaf chlorophyll fluorescence

During both laboratory and field experiments, leaf chlorophyll fluorescence parameters were also measured and calculated according to Bilger, Schreiber & Bock (1995) following the nomenclature of van Kooten & Snell (1990).

The intrinsic quantum efficiency of PSII photochemistry [(Fv/Fm) = (Fm − Fo)/Fm], where Fm is the maximal chlorophyll fluorescence yield and Fo the minimum chlorophyll fluorescence yield in the dark-acclimated state] is known to be linearly correlated with the level of PSII photo-inactivation in grapevine leaf (Hendrickson et al. 2003). At the field sites, Fv/Fm was derived from fluorescence measurements on attached leaves at 2 h intervals after 30 min of dark acclimation with leaf clips, using a portable chlorophyll fluorometer (Plant Efficiency Analyser; Hansatech, King's Lynn, Norfolk, UK).

In laboratory experiments a pulse- amplitude modulated chlorophyll fluorometer was used to determine quantum efficiency of open PSII reaction centres under irradiance [[ΦPSII = 1 − (Fs/Fm′)], where Fm′ is the maximal chlorophyll fluorescence yield during illumination and Fs is the steady-state, light-acclimated fluorescence yield], apparent electron (e) transport rate [[JPSII = (1 − (Fs/Fm′) × 0.5 × PAR × 0.85], where 0.5 is the assumed proportion of absorbed quanta used by PSII reaction centres (Melis, Spangfort & Anderson 1987), PAR is the incident irradiance and 0.85 is the assumed absorptance of grapevine leaves (Schultz 1996)] and the level of non-photochemical quenching [NPQ = (Fm/Fm′) − 1]. Leaf chlorophyll fluorescence was made simultaneously with leaf gas exchange on the same area of leaf.

Photosynthetic irradiance response at low temperatures

Using the laboratory based system described above, Riesling leaves were illuminated at a PAR of 800 µmol quanta m−2 s−1 at 25 °C until a steady-state rate of net CO2 fixation and fluorescence yield was reached at 360 µbar CO2 and 210 mbar O2 in N2 (approximately 60 min). Irradiance was then changed in step-wise increments  of  varying  magnitude from 2000 to 0 µmol quanta m−2 s−1 and measurements were made once the leaf attained a steady net CO2 fixation rate and fluorescence yield (approximately after 20 min). The response of grape leaf net CO2 fixation to irradiance was determined at 25 and 10 °C.

Photosynthetic CO2 response at low temperatures

Measurements were undertaken as for the photosynthetic irradiance response curves with the exception that once steady state was attained, ambient CO2 partial pressure in the leaf chamber (Ca) was lowered in step-wise increments of varying magnitude from 350 µbar down to 0 µbar and then increased up to 2000 µbar. The response of Riesling grape leaf net CO2 fixation to intercellular CO2 partial pressure was undertaken at 25, 20, 15, 10 and 5 °C. From these response curves, the limitation to net CO2 fixation caused by reduced stomatal conductance was estimated by calculating relative stomatal limitation (RSL) using the equation:


where ACi350 and ACa350 are the rates of net CO2 fixation at Ci = 350 µbar and Ca = 350 µbar, respectively.

Statistical analysis

The field data were organized into a randomized block design and grapevines were assigned randomly for the upper and lower sites. The results were evaluated using restricted maximum likelihood (REML) variance components analysis, which takes into account covariance, and were calculated using Genstat 4.2 (5th edition; VSN International, Oxford, UK). Growth measurements were analysed as a function of cumulative temperature and a log transformation was used to smooth curvilinearity in order to perform multiple linear regression. Analysis of variance (anova) or REML using either Origin 6.1 (Microcal Software, Northhampton, MA, USA) or Genstat 4.2 were used to evaluate variation between treatments for photosynthesis measurements. Variance of the mean of several replicates was expressed as standard error (SE). Significance probability (P) values and/or R2 values are given and results were significantly different at P < 0.05.


Photosynthesis of field-grown vines in response to microclimate

The passage of sunlight from the eastern to western halves of the sky resulted in two distinct peaks of irradiance for both halves of the north–south oriented grapevine canopy (Fig. 2). Not surprisingly, photosynthetic parameters measured on both exposed canopy halves, and to a lesser extent temperature, mimicked this trend. The diurnal change in photosynthetic and microclimatic parameters on a typical clear day showed similar trends for photosynthetic CO2 fixation (Fig. 2a) and incident irradiance (Fig. 2h) for mature, exposed Pinot Noir leaves of both east and west sides at the upper and lower microsites. Pre-dawn respiratory CO2 efflux was similar between both canopy sides and microsite. Maximum net photosynthesis (A) of east-facing leaves occurred during the morning period and was 17% higher at the upper microsite than at the lower microsite even though PAR and vpdL were very similar (Fig. 2a. e & h). However, very early morning photosynthetic rates were almost three-fold higher (P = 0.011) at the upper microsite in these leaves compared to the lower microsite, without large differences in PAR (Fig. 2a & h). Maximum A for the west-facing halves of the canopy at both upper and lower microsites were similar to the maximum A of the lower microsite east-facing leaves and to each other (P > 0.202; Fig. 2a).

Figure 2.

Photosynthetic and environmental parameters for the upper (□, ○) and lower (▪, •) site, east (□, ▪) and west (○, •) facing mature Pinot Noir grapevine leaves plotted against time on 19 December 2001 for photosynthetic carbon fixation, A (a), stomatal conductance, gs (b), intercellular CO2 partial pressure, Ci (c), transpiration, E (d), vapour pressure deficit calculated from leaf temperature, vpdL (e), leaf temperature, Tleaf (f), intrinsic light use efficiency, Fv/Fm (g) and incident light intensity, PAR (h). Pre-dawn gs and E are not presented due to condensation on the leaves. Each point is the mean ± SE of 3–4 leaves.

Stomatal conductance (gs) shows the same trend as carbon assimilation where gs is maximal in the morning and up to seven-fold higher for upper microsite east-facing leaves compared to the lower microsite (Fig. 2b). West-facing leaf gs values are similar for upper and lower microsites throughout the day with a maximum of 0.26 ± 0.05 and 0.27 ± 0.02 mol H2O m−2 s−1 in the afternoon irradiance period at both the upper and lower microsites, respectively (Fig. 2b). Pre-dawn (0500 h) measurements of gs and leaf transpiration (E) are not shown due to interference from condensation on the leaves. Interestingly, the reduced gs of lower microsite east-facing leaves was not a response to a higher vapour pressure difference (vpdL; Fig. 2e) between the intercellular air and the atmosphere. Although vpdL did increase throughout the day, there were very small differences between upper and lower microsite vpdL during the morning period when east-facing A, and gs differed.

Leaf intercellular CO2 partial pressure (Ci) declined sharply under post-dawn illumination to values between 144 and 269 µbar CO2 (Fig. 2c). Ci of west-facing leaves at both microsites followed a similar trend throughout the course of the day. During the morning peak irradiance period for the east-facing leaves, the reduction in net A for the lower microsite was associated with a reduction in gs but not with a reduction in Ci (Fig. 2a–c). The exception to this generalization was in the early morning 0600 h measurements where the average Ci of the lower microsite leaves of 171 ± 7 µbar were significantly lower than the upper microsite at 233 ± 1 µbar. By 0815 h, however, the Ci values were similar for both the upper and lower microsites (Fig. 2c).

Despite the differences in temperature and vine photosynthetic rate between the microsites, there was no difference in intrinsic photochemical efficiency of PSII (estimated by Fv/Fm) between upper and lower microsite leaves throughout the photoperiod (Fig. 2g). In fact, Fv/Fm changed very little throughout the first half of the photoperiod at either microsite, indicating that very little chilling-induced photo-inhibition took place. Pre-dawn mean Fv/Fm values were, however, lower (0.75–0.81) than values at the end of the day (0.85–0.89) when temperatures were warmer.

In summary, for the Pinot vines, the west-facing leaves showed no microsite-specific differences whereas the east-facing leaf net CO2 fixation was reduced in the early morning for the lower site. This reduction in lower site A was associated with significantly lower gs, Ci and Tleaf but not with differences in IA, vpdL or Fv/Fm. Similar diurnal measurements were carried out with Riesling vines (data not shown). Data from the Riesling vineyard were qualitatively similar to Pinot Noir, although the temperature difference was smaller between upper and lower sites (1.0–1.5 °C) and the differences in photosynthesis between the sites were correspondingly smaller (data not shown). In the early and late morning period, photosynthesis was 25 and 22% higher, respectively, for the upper east-facing leaves. Despite large differences in gs, E and vpdL the reduction in photosynthetic rate was associated with only a marginal difference in Ci, suggesting only a small stomatal component in the low temperature effect. Intrinsic quantum efficiency of PSII photochemistry, Fv/Fm, from a similar day also exhibited no variation between upper and lower site vine leaves indicating little or no effect of chilling-induced PSII photo-inactivation on the photosynthetic behaviour at the two sites.

The Pinot Noir field data from Fig. 2 and field measurements made on other days for Pinot Noir and Riesling are synthesized as plots of photosynthesis against incident irradiance for both varieties (Fig. 3). Such a treatment yields a graphical representation more similar to classic irradiance response curves for photosynthesis commonly generated for laboratory-based experiments. For the east-facing canopy of both varieties, a reduction in the mean maximum photosynthetic rate above 750 µmol quanta m−2 s−1 is observed, as described by the fitted regression lines (Fig. 3a & c). The fitted regressions exhibit a 16 and 12% reduction in light-saturated rates of carbon assimilation, Asat, for the lower microsite leaves of both Pinot Noir and Riesling varieties, respectively (Fig. 3a & c). Significantly, for Pinot Noir in particular, the regression of photosynthetic rate against light intensity for the east-facing leaves of the cooler microsite falls significantly below that for the warmer, upper site. For west-facing leaves, the photosynthetic rates for both varieties fall on the same regression at all light intensities, with little scatter in data points, regardless of microsite (Fig. 3b & d). This is because the west-facing leaf temperatures during afternoon light saturation were always at the optimum range for photosynthesis (approximately 25–28 °C according to Kriedemann 1968).

Figure 3.

A summary of photosynthetic carbon fixation, A, measured under field conditions plotted against incident irradiance, PAR, measured simultaneously for exposed Pinot Noir (a, b) and Riesling (c, d) grapevine leaves at the upper (○) and lower (▪) sites at the Tumbarumba and Canberra vineyards, respectively. Left- (a, c) and right-hand (b, d) side panels represent leaf from the east facing and west facing halves of the canopy. Each point is the mean ± SE of 3–6 leaves. The solid and dashed lines indicate fitted Gaussian regressions (R2 > 0.96) for upper and lower site leaves, respectively.

Carbon gain and growth

Total net carbon gain over a 1-d period for both exposed halves of the upper and lower microsite canopies was estimated by integration of a diurnal carbon assimilation data set and thus expressed on a per leaf area basis for both varieties (Table 1). Total (east + west) integrated net carbon gain for the upper microsite was 11 and 9% higher than the lower microsite for both Pinot Noir and Riesling, respectively (Table 1). This difference was solely due to the upper east-facing canopy fixing a total of 24 and 19% more carbon for Pinot Noir and Riesling, respectively, than its lower site counterpart (P < 0.05; Table 1). The average difference between upper and lower east-facing leaves for Pinot Noir over 4 d also showed 10.7 ± 3.7% higher net carbon gain for the upper east-facing canopy and no mean difference for the west-facing canopy. Measurements of Riesling over a period where a light wind maintained overnight air circulation exhibited no temperature difference or difference in carbon gain between microsites (data not shown). The low temperature difference would appear to be due to cold air drainage between the two microsites.

Table 1.  Total integrated carbon gain for both east- and west-facing exposed canopy halves at both upper and lower microsites over a day for both Riesling and Pinot Noir varieties
VarietyCanopy sideTotal carbon gain (mol m−2)
Lower siteUpper site
  1. Different letters indicate a significant (P < 0.05) difference between rows. Each point is the mean ± SE of three to five leaves. Data was calculated by integration of the area under carbon assimilation curves on a clear day and expressed on per leaf area basis.

Pinot NoirEast-facing leaves0.41 ± 0.01D.0.51 ± 0.00E.
West-facing leaves0.44 ± 0.01F.0.44 ± 0.02F.
RieslingEast-facing leaves0.26 ± 0.02A.0.31 ± 0.01B.
West-facing leaves0.27 ± 0.02C.0.27 ± 0.02C.

Mean total shoot lengths plotted against both time from budburst and against cumulative temperature sum are presented in Fig. 4 for the upper and lower microsite vines for both Pinot Noir and Riesling. The first vertical arrow indicates the estimated maturation date of the first source leaf (approximately 28 d after budburst; Smart & Robinson 1991) and the second arrow the beginning of flowering (60 and 78 d after budburst for Riesling and Pinot Noir, respectively). Vine shoots from both varieties at both microsites showed distinct sigmoidal growth patterns over time (Fig. 4a & c). The growth of the Riesling vine shoots was affected by a serious heat stress event (Tair > 40 °C) on 20 December 2000 that killed all apical meristems and prevented further growth with the exception of some lateral shoots.

Figure 4.

Growth of mean total shoot length (a, c) and mean natural logarithm of shoot length in mm (b, d) of upper (▪) and lower (○) site Pinot Noir (a, b) and Riesling (c, d) grapevines (Vitis vinifera L.) plotted against time from budburst (a, c) and cumulative temperature, HDD (b, d). Arrows represent the estimated time of maturation of the first source leaf and the beginning of flowering. East- and west-facing shoots were pooled for analysis. The dashed lines in (a) and (c) are  fitted  sigmoidal  regressions  that  are described as inline image and  inline image for   the upper and lower site Pinot Noir data, respectively, and inline image and inline image for the upper and lower site Riesling data, respectively.

Upper microsite growth rate, estimated as shoot length increase over time, was 63 and 34% higher than the lower microsite growth rate for Pinot Noir and Riesling varieties, respectively, between the emergence of the first source leaf and the beginning of flowering (Fig. 4a & c). Subsequent to flowering, the growth rate for both microsites declined, culminating in 34 and 43% higher total shoot lengths for the upper microsites after 110 and 107 d from budburst for Pinot Noir and Riesling, respectively (Fig. 4a & c). There was no significant difference between growth rates of east- and west-facing Riesling shoots, which were thus pooled for analysis. The Pinot Noir vines, however, showed the largest differential between the upper and lower microsites, and also showed a growth differential between east- and west-facing shoots (P < 0.001). Final mean shoot lengths for the upper microsite were 1630 ± 142 and 1326 ± 131 mm for the east- and west-facing halves of the canopy, respectively. For the lower microsite, final mean total shoot lengths were 936 ± 156 and 1158 ± 128 mm for the east- and west-facing halves of the canopy, respectively.

To determine the degree to which air temperature contributed to the above differences in growth between microsites, the data of Fig. 4a and c was re-plotted against cumulative air temperature during the duration of the experiment (Fig. 4b & d). The majority of the difference between the upper and lower microsite vine total shoot growth rates over time was removed when the data was plotted in this way, suggesting that air temperature was a major contributor to the growth differences observed. When upper and lower microsite vine shoot growth was plotted against cumulative soil temperature, it was found that these data sets did not collapse onto the same temperature relationship thus not explaining the difference in growth rates between the two microsites for either variety (data not shown). Interestingly, although final absolute shoot lengths were similar over the same time period (Fig. 4a & c), the inherent relative growth rate for Pinot Noir [0.00242 ± 0.00007 ln mm (°C day)−1] was almost one and a half times higher than the Riesling relative growth rate ([0.00161 ± 0.00004 ln mm (°C day)−1]; P < 0.001; Fig. 4b & d.

Photosynthetic response to irradiance and CO2 and low temperature

The data above strongly suggest that the microsite-dependent differences in growth observed here in the field were due to small differences in air temperature affecting the east-facing canopy. To further investigate the causal mechanisms for this effect, several experiments were conducted using glasshouse-grown material, measured under controlled conditions. Although there may be other factors influencing their response at the two field sites, it appears that photosynthetic capacity (Figs 2 & 3) and inherent relative growth rate response to temperature (Fig. 4) were more constrained for Riesling compared to Pinot Noir under similar cool climate conditions. All laboratory experiments were undertaken with the potentially more temperature-sensitive Riesling variety.

To examine the temperature response of photosynthesis in grapevine under controlled conditions, net photosynthetic carbon fixation, transpiration, stomatal conductance, intercellular CO2 partial pressure, non-photochemical chlorophyll fluorescence quenching (NPQ) and linear photosynthetic electron transport rate (JPSII) were measured at 360 µbar CO2, 210 mbar O2 in N2 and leaf temperatures of either 25 or 10 °C under varying irradiance (Fig. 5). These temperatures reflect the range commonly experienced by the east-facing canopy during illumination in the morning (Fig. 1). A major effect of the reduced temperature was a reduction in light-saturated A by approximately 42% (P < 0.001; Fig. 5a). The light-saturation point was also lower at 10 °C (800 µmol quanta m−2 s−1) compared to 25 °C (1200 µmol quanta m−2 s−1) treated leaves (Fig. 5a). Low temperature treatment also resulted in a significant decrease in both light-saturated E and gs by 80 and 49%, respectively (P < 0.001; Fig. 5b & c). Although Ci was unaffected by temperature (P = 0.2; Fig. 5d), Ci declined with increasing irradiance to stable minimum values at irradiances greater than 350 µmol quanta m−2 s−1 (Fig. 5d). At 25 °C, non-photochemical quenching (NPQ) increased in a linear fashion with increasing irradiance and did not show evidence of light saturation (Fig. 5e). By contrast, NPQ of low temperature-treated leaves exhibited a more hyperbolic relationship with irradiance, clearly saturating at much lower light intensities (Fig. 5e). The most prominent differences in NPQ due to temperature were observed between 60 and 1000 µmol quanta m−2 s−1, for example, at 530 µmol quanta m−2 s−1, 25 °C leaves developed only 23% of the NPQ at 10 °C. However, at 2000 µmol quanta m−2 s−1, the equivalent of full sunlight, NPQ values were similar at both temperatures (P = 0.24). Presumably at this point, NPQ is fully engaged at both temperatures. At 25 °C, light-saturated rates of JPSII were 2.7-fold higher than for leaves treated at 10 °C (P = 0.024; Fig. 5f). Interestingly, JPSII declined at the highest irradiance for the leaves treated at 10 °C whereas JPSII was steady in leaves treated at 25 °C.

Figure 5.

Light and temperature dependence of photosynthetic carbon fixation, A (a), transpiration, E (b), stomatal conductance, gs (c), intercellular CO2 partial pressure, Ci (d), non-photochemical quenching, NPQ (e) and linear electron transport rate, JPSII (f) for mature grapevine leaves (Vitis vinifera L. cv. Riesling). Leaves were illuminated at 25 °C (▪) or 10 °C (○) and data is plotted against photosynthetically active radiation (PAR). Vapour pressure difference at 800 µmol quanta m−2 s−1 was 0.41 ± 0.04 kPa and 0.11 ± 0.01 kPa for leaves treated at 25 and 10 °C, respectively. Average intrinsic quantum use efficiency, Fv/Fm, of the leaves was 0.80 ± 0.01. Each point is the mean ± SE of 3–5 leaves.

Net photosynthetic carbon fixation, A, Ci/Ca ratio, NPQ and JPSII were measured under controlled conditions (800 µmol quanta m−2 s−1; 210 mbar O2 in N2) at leaf temperatures between 25 and 5 °C as a function of Ci(Fig. 6). The draw down of CO2 partial pressure within the leaf was reduced at 5 °C (Fig. 6b). CO2-saturated net photosynthesis was limited by a reduction in leaf temperature with CO2-saturated rates of photosynthesis declining by 81% from 25 to 5 °C (Fig. 6a). Below 20 °C, A showed a remarkably flat relationship with CO2 partial pressure. This reduction in CO2-saturated photosynthetic capacity did not limit photosynthesis at ambient CO2 pressures until temperatures were below 15 °C (Fig. 6a). Net photosynthesis at ambient Ca, was reduced by 64% between 25 and 5 °C (Fig. 6a). The low temperature-induced reduction in photosynthetic capacity at light- and CO2-saturation was associated with a reduction in the light- and CO2-saturated rate of linear electron transport rate, Jmax, and with an increase in the level of non-photochemical quenching, NPQ (Fig. 6c & d).

Figure 6.

CO2 and temperature dependence of photosynthetic carbon fixation, A(a), intercellular to atmospheric CO2 partial pressure ratio, Ci/Ca(b), non-photochemical quenching, NPQ(c) and linear electron transport rate, JPSII(d) for mature grapevine leaves (Vitis vinifera L. cv. Riesling). Leaves were illuminated at 800 µmol quanta m−2 s−1 and measured at 25 (▪), 20 (○), 15 (▵), 10 (▿) and 5 °C (◊). Average intrinsic quantum use efficiency, Fv/Fm, of the leaves was 0.81 ± 0.01. The inset in (b) is an expansion of the linear part of the CO2 response curve in (a). Each point is the mean ± SE of 3 leaves.

A summary of the temperature response of light-saturated assimilation at either ambient or elevated CO2 partial pressure is presented in Table 2. Relative stomatal limitation, RSL (calculated as the difference between A at Ca and Ci of equivalent value; see Eqn 2), was relatively constant between 25 and 15 °C, at 20.6–22.5%, and was dramatically reduced to 0% below 15 °C (Table 2). Carboxylation efficiency (CE), calculated from the initial slope of the photosynthetic CO2 response curve under CO2-limiting conditions also declined as a function of temperature (P < 0.05; Table 2). CE was reduced by 71% between 25 and 5 °C (Table 2). Upon lowering of the oxygen partial pressure from 210–20 mbar O2, photosynthetic CO2 uptake was stimulated under warm conditions and this sensitivity was reduced and even partially reversed at lower leaf temperatures (data not shown).

Table 2.  Summary of the temperature response of photosynthetic parameters from Fig. 6a at ambient (Ca = 360 µbar) and high (Ca = 1950 µbar) CO2 partial pressures
Leaf temperature (°C)252015105
  1. Each point is the mean ± SE of 3–5 leaves. RSL is the relative stomatal limitation, Asat and Amax are the light-saturated rates of photosynthesis at ambient and saturating CO2 partial pressures, respectively, and CE is the leaf carboxylation efficiency calculated from the initial slope of the photosynthetic CO2 response curve.

Relative stomatal limitation (RSL: %)   21 ± 3   21 ± 2   23 ± 4   13 ± 4    1 ± 3
Amax (µmol CO2 m−2 s−1;Ca = 1950 µbar) 28.9 ± 0.1 22.0 ± 0.1 14.9 ± 0.1 10.5 ± 0.1  5.4 ± 0.1
Asat (µmol CO2 m−2 s−1; Ca = 360 µbar) 14.4 ± 0.1 13.7 ± 0.1 11.6 ± 0.1  9.3 ± 0.7  5.2 ± 0.4
CE (µmol m−2 s−1 µbar−1)0.073 ± 0.0010.063 ± 0.0020.054 ± 0.0030.038 ± 0.0050.021 ± 0.001


Growth at low temperatures in the field

Large average temperature differences and/or severe low temperature treatments typically inhibit both grapevine and kiwifruit vine growth (Buttrose 1969; Greer, Cirillo & Norling 2003). The long-term meteorological and short-term temperature data show that field-grown grapevines in this study were exposed to frequent chilling events and wide diurnal temperature ranges (Fig. 1). Not surprisingly, growth of field-grown grapevine shoots was found to respond directly to temperature and this growth response to temperature. This has been well characterized for other crop species such as wheat (Sayed 1995), cotton (Roussopoulos, Liakatas & Whittington 1998) and kiwi vine (Greer et al. 2003). The interesting result of this present study, however, is that the growth data for both grapevine varieties suggests that the large mid-season growth differential between vines upon a slope at an upper (warmer) and lower (cooler) microsite is the result of cumulative effects of a small but consistent, 1–3 °C, air temperature difference over the preceding period (Fig. 4). If this is the case then in the long term the 1 and 3 °C warmer upper microsites for both varieties, where the extent of low temperature excursions was more limited, had a slight growth advantage over the cooler lower microsites. Similar observations, although over a much larger temperature range, have previously been reported in the literature for other crop plants (Sayed 1995). The results of this study are also qualitatively similar to the study by Blennow et al. (1998) in which an average minimum temperature difference of 2 °C severely affected growth resulting in an 80% reduction in biomass accumulation of snowgum seedlings in the colder sites during a spring growth period. Unfortunately, harvesting of grapevine shoots midseason for dry weights was impractical in the commercial vineyards used in this study and so further studies of this phenomenon should include such supporting evidence.

Measurements of soil temperature were unable to account for the difference in growth rates between the two microsites for either grapevine variety. It is clear from the data presented here that grape vine shoot growth showed a strong response to quite small differences in ambient air temperature. Two explanations for this observation are likely. First, shoot growth itself may respond directly to temperature through an effect on cell division and expansion (see Grace 1988). Second, the provision of carbon for growth may be impaired by low temperature (by a direct effect on either photosynthesis or respiration). In this context, at the Pinot Noir vineyard where the temperature differences between upper and lower microsites were larger (3 °C), there was a significant difference between east- and west-facing shoot length at both microsites despite the fact that both canopy sides experienced similar soil and air temperatures throughout the day (Fig. 1). The only difference between the two halves of the canopy in these experiments was the temperature at which they received illumination. This result suggests that a direct effect of temperature on cell growth or respiration is unlikely to explain the growth differences seen here and implies that light-dependent carbon acquisition may be of more importance. It is also of note that the divergence in shoot length between microsites for both varieties occurred subsequent to the emergence of source leaves (Smart & Robinson 1991). This indicates that canopy leaves at full photosynthetic capacity are required to realize the temperature-induced variations in growth rate. In young pot-grown vines, net photosynthesis and total dry mass accumulation is well correlated. In the study by Ferrini et al. (1995), low temperature (20 °C) grown vines had lower shoot lengths and lower total dry weights after an identical growth period than the warmer (27 °C) conditioned vines. However, it is interesting that for the mature vines used in our field study, containing considerable starch reserves in the wood and elsewhere, shoot elongation remained closely linked to the provision of de novo photosynthate in the early growth season.

Photosynthesis at low temperatures in the field

Photosynthesis measurements made here on mature, exposed grapevine leaves demonstrated that light-saturated net carbon fixation was indeed reduced by low morning temperatures (Figs 2, 3, 5 & 6). Effects of early morning chilling on crop yield, manifested through photo-inactivation of PSII, have been well characterized both in the laboratory and in the field (Long, Humphries & Falkowski 1994). However, sustained photo-inactivation of the PSII reaction centre at low temperature and high irradiance played little role in this reduction (Fig. 2). The integration of the area under the assimilation curves for a typical cool clear day demonstrated that, at light-saturation, the east-facing Pinot Noir leaves at the marginally warmer upper site contributed to a daily average of 10% higher carbon gain. The observation that the difference in carbon gain and growth between the microsites is solely driven by the east-facing canopy is also highly supportive of the hypothesis that an effect of low temperature on photosynthesis is the major contributor to the microsite growth differences observed (Table 1). Together with the observation that photo-inactivation does not occur in these field experiments and that despite contrasting temperatures pre-dawn leaf respiration was similar across canopy aspect and microsite (Fig. 2a), these data suggest that the low temperature effects on photosynthesis and growth seen here could be a result solely of an innate response of photosynthesis to temperature.

While the canopy and microsite data discussed above suggests that respiration may not play a role in the growth responses observed here, its contribution cannot be ruled out. At low irradiance, respiration is the most influential factor affecting net carbon gain (Loveys et al. 2002). In this present field study, measurements of light-saturated net photosynthesis were made on the exposed leaf canopy only. Under these conditions, respiration would account for a very small proportion of leaf photosynthesis confirming a temperature-induced contribution of photosynthesis to the net carbon gain and to the shoot growth of field-grown vines. However, considering the close relationship between leaf respiration and net carbon gain under low-light conditions, to accurately quantify the contribution of net photosynthesis to ‘net’ carbon gain and growth, analysis of either shade leaves within the canopy or whole canopy photosynthesis at low temperatures should be undertaken. In this context, from the beginning to the end of flowering, when the field photosynthesis measurements were made, Downton & Grant (1992) estimate that approximately 60% of the leaf area of each vine is exposed. This large proportion of canopy exposure implies that a marginally higher, early morning carbon gain for upper site east-facing vines would have a substantial effect when extrapolated to the majority of the canopy. Indeed these experiments suggest that temperature dependence of vine photosynthetic carbon gain could be very important during canopy development in the early growth season when the majority of source leaves are light saturated.

The direct comparison between daily carbon gain of both canopy sides revealed that the west-facing leaves yielded less total carbon (Table 1). The reduced carbon gain was principally associated with limited maximum stomatal conductance for the west-facing leaves and was a response to considerably higher vapour pressure deficit in the afternoon rather than extreme leaf temperature (Fig. 2). This phenomenon was exaggerated for upper site vines where east-facing photosynthetic capacity was less constrained by low temperatures. However, the growth data tends to suggest that the temperature effect overrides this potentially mitigating factor in the long term (Fig. 4).

Low temperature limits grapevine photosynthesis via several mechanisms

The reduction in photosynthesis at low temperatures in field-grown vines measured in this study could have been due to a number of factors. Stressful conditions can impair photosynthesis via both stomatal and non-stomatal processes in grapevine as demonstrated for grapevines experiencing drought stress (Maroco et al. 2002). Field experiments indicated that there was both a stomatal and non-stomatal component to the low temperature-induced depression of carbon fixation in the morning (Fig. 2). This occurred in both varieties (Fig. 4); however, the focus of the glasshouse experiments was the potentially more temperature sensitive variety Riesling (as demonstrated in Fig. 4b & d). Although the glasshouse-grown material was not ‘cold-acclimated’ as for the field-grown material, growth of vines under controlled conditions and photosynthetic measurements in the laboratory allowed further quantification of the component processes contributing to the low temperature limitation of grapevine photosynthesis (Figs 5 and 6).

Stomatal limitation of photosynthesis

This study has demonstrated that field- and glasshouse-grown grapevines exhibit stomatal closure in response to low temperatures independent of CO2 partial pressure or vapour pressure difference. A significant stomatal component in the low temperature response of photosynthesis in Pinot Noir leaves in the field was demonstrated in Fig. 2, where photosynthesis was clearly Ci-limited for the east-facing lower site leaves experiencing saturating irradiance at temperatures below 10 °C in the early morning. The low temperature-induced stomatal limitation is supported by the lack of significant variation in either vpdL or irradiance that might otherwise explain stomatal function in the early morning period (Fig. 2). Experiments with non-water-stressed, glasshouse-grown leaves subjected to low temperature treatment at approximately constant Ci exhibited reduced gs (Figs 5 & 6). This result confirms the observations of Flexas et al. (1999) where Riesling grapevines chilled in the dark followed by a brief period of illumination resulted in a reduction of light-saturated gs in the absence of water deficit. In this study, vapour pressure difference can be ruled out as a factor in cold-induced stomatal closure since vpdL decreased with decreasing leaf temperature and was maintained well below 0.5 kPa during laboratory experiments. Wilkinson, Clephan & Davies (2001) have proposed that cold-induced stomatal closure is a response to increased apoplastic calcium uptake of guard cells rather than a drought-induced abscisic acid response. However, increases in leaf ABA contents are believed to be responsible for prolonging this cold-induced stomatal closure (Pérez de Juan, Irigoyen & Sanchez-Dias 1997). Earlier work by Liu et al. (1978) also supports this hypothesis for grapevine. Liu et al. (1978) reported partial stomatal closure independent of leaf water potential for light-saturated and cold-acclimated Vitis lambruscana leaves subsequent to nights with minimum temperatures below 10 °C.

Non-stomatal limitation of photosynthesis

In the early morning period the change in Ci was not as proportionately great as that of A for the lower east-facing leaves (Fig. 2). Later the same morning, when Ci was similar between upper and lower site east-facing leaves the reduced A of lower site leaves persisted suggesting some form of non-stomatal limitation. Similar to the field observations, the treatment of the glasshouse-grown vines at 10 °C, induced a reduction in light-saturated A as well as gs but in this case the reduction in gs did not limit the partial pressure of CO2 in the intercellular air space as it did in field conditions (Fig. 6). By determining the response of photosynthesis to varying CO2 concentration, it was possible to quantify the stomatal contribution to restriction of photosynthesis at low temperature. This parameter, called relative stomatal limitation (RSL), was similar (in the temperature range 15–25 °C) to the 21.7 ± 1.7% (Maroco et al. 2002) and the approximately 25% (Escalona, Flexas & Medrano 1999) reported for field-grown irrigated vines from a warm climate. In this study, RSL remained steady (approximately 20%) above 15 °C but indicated that below 15 °C, stomatal closure was a response to limited photosynthetic capacity rather than the converse (Table 2). A similar effect was observed in several native and cultivated species acclimated to field conditions where RSL declined to between 0 and 8% at 10 °C, a value similar to that reported for the warm-acclimated vines of this present study (Sage & Sharkey 1987). Thus in addition to a reduction in gs at saturating irradiance and low temperature, there is a potential non-stomatal limitation of photosynthesis contributing to reduced carbon gain for early morning east-facing leaves (Table 2).

A clue to the nature of this non-stomatal limitation can be found in the observation that light- and CO2-saturated photosynthetic capacity (Amax) in glasshouse-grown Riesling was greatly reduced at all temperatures below 25 °C (Table 2). This phenomenon has previously been reported to be the result of a reduced rate of RuBP regeneration. RuBP regeneration becomes limited at low temperatures due to one of three reasons: (1) limitation of the rate at which light-harvesting and electron transport produce ATP and NADPH; (2) limitation of the rate at which the stromal bisphophatases regenerate RuBP in the photosynthetic carbon reduction cycle; or (3) restriction of the rate at which end-product synthesis consumes triose-phosphates and regenerates inorganic phosphate (Pi) for photophosphorylation (Allen & Ort 2001). RuBP regeneration, represented in Table 2 as Amax, closely matched grapevine photosynthesis at ambient CO2 partial pressures (Asat) below 15 °C (Table 2). Amax in barley, tomato, capsicum, poplar and figwort leaves have also been reported to be similarly affected by low temperatures and was associated with O2-insensitivity (Sage & Sharkey 1987; Labate & Leegood 1988), an indicator of end-product limitation of photosynthesis (Leegood & Furbank 1986; Sharkey et al. 1986).

The extremely flat nature of the photosynthetic CO2 response curve at high CO2 partial pressures can be interpreted as evidence for end-product limitation of photosynthesis in grapevines (Harley & Sharkey 1992). O2-insensitivity was observed for grapevine photosynthesis at low temperatures (data not shown). This evidence, in addition to the higher NPQ and lower JPSII, strongly suggests that low temperature limits RuBP regeneration for warm-acclimated grapevines via end-product limitation of phosphate-recycling and photophosphorylation. Interestingly, at low temperature, JPSII declined at high Ci without any apparent effect on A (Fig. 6). This observation suggests that at low temperature and high CO2 there is a repartitioning of electrons from one sink to another, not reflected in the CO2 assimilation rate and that the capacity of electron transport is in excess of that required for carbon assimilation.

It should also be noted that cold acclimation generally leads to increased photosynthetic rates at all low temperatures and a downward shift in the photosynthetic optimum leaf temperature (Badger, Björkmann & Armond 1982). Wide spring diurnal temperature ranges may ensure that grapevines are acclimated to such a broad temperature range that regular low temperature excursions may still cause end-product limitation. The similarity of the data between this study and that of Sage & Sharkey (1987), which was undertaken on field-grown tomato, capsicum, poplar and figwort plants infer that low leaf temperature limitation of photosynthesis by end-product limitation may still be relevant to diurnal grapevine carbon gain, particularly during periods of the day when temperatures are relatively low enough.

Apart from feedback effects of low temperature limiting carbon metabolism on photosynthetic processes in general, there may also be a direct inhibitory effect of temperature on thylakoid flux (Figs 5 & 6). A study by Ott et al. (1999) demonstrated a direct limitation of electron transport capacity by low temperature and invoked diffusional limitations on plastoquinone and plastocyanin. However, any decrease in electron transport capacity at this point under saturating light, would feedback on the electron transport chain causing reduction of QA, the primary quinone electron acceptor of PSII, and likely lead to photo-inactivation of the PSII reaction centre. The general absence of chronic photo-inactivation of field-grown vines in this study (Fig. 2g) and other studies (Chaumont et al. 1997) implies a generally low reduction state of QA suggesting that direct temperature effects on electron transport are not a major contributing factor in the present study. Incidentally, it is interesting to note that the ratio of electron flux calculated from fluorescence to electron flux required for CO2 assimilation declines markedly with temperature, indicating a reduction in available alternative electron acceptors.

In addition to the reduction of photosynthetic capacity by limited RuBP regeneration, the initial slope of the CO2 response curves (Table 2) indicated that carboxylation efficiency was reduced at all temperatures below 25 °C and particularly at 5 °C in grapevine. The slope of this relationship is often called the Rubisco-limited rate of photosynthesis because of its direct dependence on the maximum activity of Rubisco (von Caemmerer et al. 1994). There will be a direct effect of temperature on Rubisco catalytic efficiency (Sage 2002), but Rubisco activation state also decreases when end-product limitation depletes the pool of inorganic phosphate and the concentration of RuBP becomes suboptimal (Sharkey 1990). It is unlikely that the total amount of Rubisco present was affected under short low temperature treatments. These data therefore suggest that, in addition to the catalytic turnover rate of RuBP carboxylation, Rubisco activation state may be impaired by low temperature-induced end-product limitation in grapevine leaves.

In conclusion, the results of this study indicate that the accumulation of small differences in mean minimum air temperatures within two cool climate vineyards resulted in a large variation in grapevine shoot growth over time. The greater accumulated carbon gain by the exposed vine canopy at marginally warmer (1–3 °C) microsites was associated with more marginally higher temperature conditions in the cool early morning period for exposed east-facing leaves. We suggest that this was at least in part due to low temperature effects on photosynthesis through both stomatal and non-stomatal mechanisms unrelated to photo-inactivation of PSII by chilling. This low temperature phenomenon, although associated with stomatal restriction of intercellular CO2 partial pressure and Rubisco inactivation above 15 °C was more likely a response to RuBP regeneration and further Rubisco inactivation below 15 °C.


We gratefully acknowledge the field support provided by Jack Egerton and Wayne Pippen and Dr Adrienne Nicotra for her provision of equipment. We thank Bill Parker of Stringybark Hill vineyard and Bruce Guthrie of Tumbarumba Southcorp Wines for allowing unrestricted access to their respective vineyards. We thank Adj. Professor Paul Kriedemann for his expert advice on the preparation of this manuscript. L.H. is indebted to Professor Barry Osmond who established the ARC Strategic Partnerships with Industry-Research and Training Scheme Grant (C19906986) that funded this study.