The magnitude of the response of stomatal conductance to a change in the concentration of carbon dioxide external to the leaf from 350 to 700 cm3 m–3 was found to be extremely variable from day to day in the field in Glycine max, Hordeum vulgare and Triticum aestivum. It was found that the leaf-to-air water vapour pressure difference (LAVPD) during the midday measurements of the stomatal response to carbon dioxide affected the magnitude of the response. On days when LAVPD was low, no significant change in conductance occurred with the increase in carbon dioxide concentration. When LAVPD was higher, conductance decreased by 24–52% with the increase in carbon dioxide within a few minutes. The sensitivity of conductance was approximately linearly related to LAVPD in wheat and barley. Experiments with G. max in the field indicated that, on days with low LAVPD, increasing the LAVPD just around the measured portion of a leaflet made stomatal conductance responsive to increased carbon dioxide. This result was also obtained under laboratory conditions with G. max, Helianthus annuus and Amaranthus retroflexus. In G. max, it was determined that leaves in which conductance was not responsive to the increase in carbon dioxide could be made responsive even at low LAVPD by the injection of abscisic acid into their petioles. Because it is known that abscisic acid sensitizes stomata to carbon dioxide, these results are consistent with the idea that abscisic acid may be involved in the response of stomatal conductance to changes in LAVPD.
The difference in the partial pressure of water vapour in air inside and outside leaves, the leaf to air water vapour pressure difference (LAVPD), is one of the most important variables affecting leaf stomatal conductance. The sequence of events causing responses of stomatal conductance to changes in LAVPD is uncertain. Because LAVPD is a measure of the driving force for transpiration (Cowan 1977), and because stomata are turgor-operated valves, investigations of the response of conductance to LAVPD have focused on leaf water relations. Monteith (1995) suggested that responses not associated with ‘patchy’ stomatal closure were consistent with the reduction of stomatal conductance with increased LAVPD being caused by higher transpiration. However, this explanation is not always satisfactory (Comstock & Ehleringer 1993; Saliendera et al. 1995; Bunce 1997). Because many stomatal responses previously attributed to changes in leaf water status have been shown to be mediated by abscisic acid (cf. Tardieu et al. 1996), there has been renewed interest in the hypothesis that stomatal responses to LAVPD could be mediated by abscisic acid (Hall & Kaufman 1975; Grantz & Schwartz 1988). It was recently reported that very low carbon dioxide concentrations eliminated the response of conductance to changes in LAVPD in three species which responded normally at higher carbon dioxide concentrations (Bunce 1997). This suggested that abscisic acid might mediate the response of conductance to LAVPD, since it is known that carbon dioxide is necessary for stomatal responses to abscisic acid (e. g. Raschke 1975). It was suggested (Bunce 1997) that LAVPD could affect the rate of delivery of abscisic acid to the guard cells, and thereby influence stomatal conductance.
The present paper provides a different type of correlation which is also consistent with the idea that abscisic acid may be involved in the response of stomatal conductance to changes in LAVPD. These experiments arose from attempts to understand why the short-term response of stomatal conductance to increases in carbon dioxide concentration varied greatly from day to day in the field.
MATERIALS AND METHODS
Measurements were made on leaves of Hordeum vulgare L. cv. Wyson, Triticum aestivum L. cv. Coker and Glycine max L. Merr. cv. Clark grown outdoors in field plots in 1995 and 1996 at Beltsville, MD, and also on G. max, Helianthus annuus L. cv. Mammoth, and Amaranthus retroflexus L. grown in an air-conditioned glasshouse. Day/night temperatures in the glasshouse were controlled at 28/19 ºC each ± 2 ºC, and carbon dioxide was injected to maintain daytime values > 350 cm3 m–3. A blower constantly circulated glasshouse air through a heat exchanger, and provided an air speed of about 0·5 m s–1 over leaves, and produced gentle leaf fluttering. The lower temperature was maintained for 12 h beginning 8 h after solar noon. In the field, seeds of H. vulgare and T. aestivum were planted directly in a fertile silt-loam soil in early October 1994 and 1995, and measurements of stomatal conductance were conducted the following springs. Seeds of G. max were planted in June of 1995 and 1996 at the same field site, and measurements were made in July and August of those years. In the glasshouse, plants were grown in the summer of 1996. Plants were rooted in vermiculite in 20-cm-diameter pots which were flushed daily with a complete nutrient solution.
Stomatal conductance to water vapour of leaves was measured using a CIRAS-1 portable photosynthesis system (PP Systems, Haverhill, MA), equipped with automatic carbon dioxide control and a broad-leaf chamber with an infrared leaf temperature sensor. The leaf chamber window contained an infrared filter, and the chamber had ventilated internal and external heat exchange surfaces. Both in the field and in the growth cabinet, air temperatures in the leaf chamber and leaf temperatures were within 3 ºC of the temperature of the outside air, even at high irradiance. Measurements were made on fully expanded upper leaves. LAVPD was calculated from the ratio of transpiration to stomatal conductance computed by the CIRAS system, and is referenced to the water vapour pressure of air inside the leaf boundary layer. Leaf sections (2·5 cm2) were illuminated either with sunlight in the field, or with a halogen lamp in laboratory measurements on plants grown in the glasshouse. For the grass leaves, if leaf width was insufficient to cover the cuvette window, two leaves placed edge to edge were used. The temperature of the leaf chamber was uncontrolled in the field, but was controlled by placing the entire system in a temperature-controlled growth cabinet for the laboratory measurements. Leaves were first equilibrated at an external carbon dioxide concentration of 350 ± 10 cm3 m–3, and then the external concentration was changed to 700 ± 10 cm3 m–3. Stomatal conductance was monitored for periods of up to 30 min following the change in carbon dioxide concentration. In the field, measurements were made near midday on clear days, and PPFD on the measured section of leaf exceeded 1·5 mmol m–2 s–1. Stomatal responses of six replicate leaves per species were measured on each occasion. During measurements in the laboratory on plants grown in the glasshouse, the plants and instruments were at 25 ºC, 1·0 mmol m–2 s–1 PPFD, 18 ºC dew point temperature and 350 cm3 m–3 carbon dioxide concentration, while the measured leaf section was at a PPFD of 2·0 mmol m–2 s–1.
In the laboratory measurements and some of the field measurements, the water content of the air was manipulated to alter LAVPD. The response of stomatal conductance to the increase in carbon dioxide concentration was first measured at a low value (mean = 0·8 kPa) of LAVPD. The same leaf section was then returned to the lower carbon dioxide concentration until conductance returned to the original value. The water content of the air just around the measured section of leaf was then reduced by drying the air entering the leaf chamber. The net result was a higher value of LAVPD (mean = 1·6 kPa), with decreases in leaf temperature of 0·4–0·8 ºC. The area of the measured leaf section (2·5 cm2) was small compared with the total area of the measured leaf (80–150 cm2), and with the total plant leaf area (> 1000 cm2) in all species. The stomatal conductance reached a stable value at the higher LAVPD within 10 min. After about 20 min at the higher LAVPD, the carbon dioxide concentration was increased, and the stomatal response monitored. Condensation of water vapour inside the cuvette or in the air line leading out of the cuvette can occur during measurements at low LAVPD, and would cause errors in estimates of stomatal conductance. We carefully checked for this error by confirming that removing a leaf from the cuvette immediately resulted in a transpiration rate of zero, indicating that no water had condensed in the system.
In G. max and H. annuus leaves grown in the glasshouse, treatment with abscisic acid was also an experimental variable. Stomatal response to the carbon dioxide increase (see earlier) was first measured at low (0·9–1·0 kPa) LAVPD. The carbon dioxide concentration was then returned to the lower value, and 50 mm3 of a 1·0 mol m–3 solution of ± ABA in 1% ethanol was injected into the petiole of the measured leaf. The upper surface of the petiole was punctured with a 0·7-mm-diameter hollow needle, and the solution ejected from the needle. Any solution not absorbed immediately was taken up in the transpiration stream within 2 min. After about 30 min, when stomatal conductance had reached a new, lower value and had been stable for about 10 min (see ‘Results’ for a representative time-course), the carbon dioxide concentration was raised, and the stomatal response monitored. Control injections of 1% ethanol solutions did not cause changes in stomatal conductance in G. max leaves, but they caused rapid complete stomatal closure in H. annuus leaves. Therefore no further experiments were done with H. annuus leaves.
In the field, when stomatal conductance was measured at high PPFD under essentially ambient temperature and humidity conditions, the response of stomatal conductance to the imposed doubling of carbon dioxide was extremely variable from day to day for all three species. On some days there was no significant response of conductance, while on other days conductance decreased to as little as 0·36 of the value at the lower concentration (Fig. 1, Table 1). On the days in which the LAVPD was less than 1·2 kPa, there was no significant response of conductance to increased carbon dioxide, while a response was always detectable on days with higher LAVPD (Fig. 1, Table 1). While LAVPD and temperature were roughly correlated in the field for the wheat and barley data, the temperature range over which there was little change in conductance with increased carbon dioxide broadly overlapped the range where conductance was reduced (Fig. 2). In soybeans, there was no consistent difference in leaf temperature between days when conductance varied with carbon dioxide and days when it did not (Table 1).
Table 1. . The ratio of stomatal conductance (g) at 700 cm3 m–3 to that at 350 cm3 m–3 carbon dioxide concentration on days with lower and higher leaf to air water vapour pressure deficit (LAVPD) for G. max under field conditions in 1995. Stomatal conductance was first measured at 350 cm3 m–3 carbon dioxide concentration and then the concentration was increased to 700 cm3 m–3. Responses were measured near midday on clear days, with leaves at the ambient temperature and water vapour conditions. When stomatal conductance decreased at higher carbon dioxide concentration, leaf temperature changed by no more than 1°C, and LAVPD increased by less than 0·2 kPa. LAVPD and leaf temperature in the table refer to values before conductance changed with carbon dioxide
Tests showed that, on days when there was no change in conductance with carbon dioxide, there was no change for up to 30 min, whereas on days when conductance did respond, the change was complete within a few minutes. On days with low LAVPD and no response of stomatal conductance to increased carbon dioxide, conductance responded within a few minutes to changes in other environmental factors, such as PPFD or LAVPD. For example, in G. max on 16 August 1996 (i.e. the ‘field’ data in Table 3), at low LAVPD, conductance decreased by 35% within 1 min of decreasing PPFD to 100 μmol m–2 s–1, and decreased by 40% within 2 min of drying the air entering the cuvette, whereas no change in conductance with increased carbon dioxide occurred for up to 10 min. These results led to the hypothesis that, at low LAVPD, stomatal conductance did not respond to the increase in carbon dioxide. Because the whole plant and the measured leaf were at essentially the same LAVPD, the LAVPD dependence could be a either a localized leaf response or a response based on the whole-plant transpiration rate.
Table 3. . Effect of injection of 50 nmol of ±ABA on responses of stomatal conductance (g) to change in external carbon dioxide concentration from 350 to 700 cm3 m–3. Stomatal responses of leaves of G. max were measured at 1·0 mmol m–2 s–1 PPFD, a leaf temperature of 26–28 °C, and a leaf to air water vapour pressure deficit of 0·9–1·1 (low) or 1·8–2·0 (high) kPa
The hypothesis that low LAVPD caused the lack of stomatal response to carbon dioxide increase was tested directly in the field with G. max, and in the laboratory with G. max, H. annuus and A. retroflexus grown in the glasshouse. These experiments also served to determine whether the LAVPD dependence occurred at the leaf or the whole-plant level. On a day with low LAVPD in the field, and no response of conductance to carbon dioxide in soybean, decreased conductance with increasing carbon dioxide occurred in each of four leaves at higher LAVPD (Table 2). The same result was found in laboratory measurements on G. max, H. annuus and A. retroflexus grown in the glasshouse (Table 2).
Table 2. . Effect of leaf to air water vapour pressure deficit (LAVPD) on responses of stomatal conductance to change in external carbon dioxide concentration from 350 to 700 cm3 m–3. Stomatal response to carbon dioxide was measured at low LAVPD, then the LAVPD around the measured section of leaf was increased and the response to change in carbon dioxide measured again. PPFD was 2·0 mmol m–2 s–1 in the laboratory and > 1·5 mmol m–2 s–1 in the field. *indicates a significant effect of measurement carbon dioxide on stomatal conductance within a LAVPD treatment at P = 0·05, using paired t-tests. LAVPD is in kPa and conductance (g) is mmol H2O m–2 s–1. Values are ± SE, and n indicates the number of leaves measured. When stomatal conductance decreased at higher carbon dioxide, leaf temperature changed by less than 1°C, and LAVPD increased by less than 0·3 kPa. LAVPD in the table refers to that before conductance changed with carbon dioxide
In G. max, a decrease in stomatal conductance with increasing carbon dioxide even at low LAVPD was induced by injection of ABA into the petiole while the leaf was at the lower carbon dioxide concentration (Fig. 3, Table 3).
Insensitivity of stomatal conductance to a doubling of the external carbon dioxide concentration from 350 to 700 cm3 m–3 occurred in all five species examined when leaves were measured at low LAVPD. Experiments on three of the species indicated that exposure of just the measured portions of leaves to higher LAVPD quickly restored more typical stomatal sensitivity, as was suggested by the patterns of day-to-day variation in sensitivity in the field. The LAVPDs for which conductance was insensitive to carbon dioxide were lower than typically used in laboratory gas exchange systems. The data for wheat and barley suggest a progressive increase in sensitivity of conductance to carbon dioxide as LAVPD increased, similar to that reported for sesame (Hall & Kaufman 1975). In our field data, temperature was not constant from day to day, and interactions between temperature and the LAVPD dependence of carbon dioxide sensitivity may have contributed to the scatter in the relationship. However, the experiments in which the water content of the air was manipulated to alter LAVPD involved changes in leaf temperature of less than 1 ºC, and indicated that LAVPD, not leaf temperature, was the primary factor affecting stomatal response to carbon dioxide. It does not necessarily follow that all cases of stomatal insensitivity to carbon dioxide can be reversed by high LAVPD.
In our data, when stomatal conductance increased in response to the increase in carbon dioxide, leaf temperature and LAVPD were also altered slightly, so the net changes in conductance cannot be solely attributed to carbon dioxide. However, the initiation of the change in conductance can be attributed to the carbon dioxide treatment, as neither leaf temperature nor LAVPD varied with carbon dioxide concentration until leaf conductance responded.
It is known that abscisic acid sensitizes stomata to carbon dioxide (e. g. Dubbe et al. 1978), and the sensitization of stomata to carbon dioxide by elevated LAVPD found here suggests that increased LAVPD may increase abscisic acid concentration. Grantz & Schwartz (1988) noted that the facts that the initial and final responses of stomatal conductance to changes in LAVPD are opposite in sign, and that the time courses of stomatal responses to LAVPD and to PPFD are similar, suggest the involvement of abscisic acid in stomatal responses to LAVPD. Tardieu et al. (1996) found that stomatal conductance in sunflower was correlated with xylem abscisic acid concentration over a range of conditions, including a range of VPDs, although it is unclear from their data whether VPD affected xylem abscisic acid concentration separately from its effect on water potential.
Ward & Bunce (1986) found that an increase in VPD at one surface of sunflower leaves reduced the stomatal conductance of the opposite surface. The effect could not be attributed to a change in leaf water status, and it was suggested that there was a signal, possibly abscisic acid, produced by high LAVPD, which was transported to the opposite leaf surface.
The fact that insensitivity of stomatal conductance to increased carbon dioxide caused by low LAVPD in soybean was reversed by the application of abscisic acid in the present experiments is also consistent with the idea that the increase in stomatal sensitivity to carbon dioxide with increased LAVPD may involve an increase in abscisic acid with the increased LAVPD. This idea is further supported by the observation that no stomatal response to change in LAVPD occurred at very low carbon dioxide concentrations (Bunce 1997), when response to abscisic acid is minimal (e. g. Raschke 1975). Lack of LAVPD sensitivity at low carbon dioxide was also reported for sesame (Hall & Kaufman 1975). In total, several different types of correlations suggest the possible involvement of abscisic acid in responses of stomatal conductance to changes in LAVPD, although direct evidence is lacking.
Regardless of the mechanism, the data presented here indicate that the magnitude of the response of stomatal conductance to changes in carbon dioxide concentration varies with humidity, and that the humidity effect can be quite localized.