Distinct light responses of the adaxial and abaxial stomata in intact leaves of Helianthus annuus L.

Authors


Y. Wang. Fax: +81 3 5841 4465; e-mail: wangy@biol.s.u-tokyo.ac.jp

ABSTRACT

Using a laboratory-constructed system that can measure the gas exchange rates of two leaf surfaces separately, the light responses of the adaxial and abaxial stomata in intact leaves of sunflower (Helianthus annuus L.) were investigated, keeping the intercellular CO2 concentration (Ci) at 300 µL L−1. When evenly illuminating both sides of the leaf, the stomatal conductance (gs) of the abaxial surface was higher than that of the adaxial surface at any light intensity. When each surface of the leaf was illuminated separately, both the adaxial and abaxial stomata were more sensitive to the light transmitted through the leaf (self-transmitted light) than to direct illumination. Relationships between the whole leaf photosynthetic rate (An) and the gs for each side highlighted a strong dependence of stomatal opening on mesophyll photosynthesis. Light transmitted through another leaf was more effective than the direct white light for the abaxial stomata, but not for the adaxial stomata. Moreover, green monochromatic light induced an opening of the abaxial stomata, but not of the adaxial stomata. As the proportion of blue light in the transmitted light is less than that in the white light, there may be some uncharacterized light responses, which are responsible for the opening of the abaxial stomata by the transmitted, green light.

Abbreviations
AB

abaxial epidermis

AD

adaxial epidermis

An

rate of photosynthesis

Ci

intercellular CO2 concentration

gs

stomatal conductance

gs*

stomatal conductance per stoma

PPFDs

photosynthetically active photon flux densities

VPD

vapour pressure deficit between the air and the leaf

INTRODUCTION

Stomata provide the major pathway for the exchange of CO2, O2 and water vapour between the atmosphere and the plant. Stomata respond to many environmental variables, opening in response to the low CO2 concentration, high light and high humidity, and closing in response to the high CO2 concentration, low light or darkness, and low water availability. In nature, these factors always exert compound effects on stomatal movement (Sharkey & Raschke 1981a,b; Zeiger & Zhu 1998; Talbott, Rahveh & Zeiger 2003a). To investigate in vivo responses of stomata to light, separately from those to the other factors, therefore, we have to precisely control the other environmental factors.

The light response of stomata has at least two components (Sharkey & Raschke 1981a,b), the blue-light-mediated response and the photosynthesis-mediated response. The blue-light response has been shown to involve light activation of the plasma membrane H+-ATPases (Kinoshita & Shimazaki 1999). The most likely receptors for this response to be phototropins (Kinoshita et al. 2001; Doi et al. 2004) and/or zeaxanthin (Zeiger & Zhu 1998; Talbott et al. 2003b, 2006). Photosynthesis-dependent responses involve photosynthesis in stomatal guard cells as well as that in mesophyll cells (Zeiger et al. 2002). Vigorous mesophyll photosynthesis lowers the Ci, and the Ci thus lowered causes stomatal opening (Scarth 1932; Sharkey & Raschke 1981a; Mott 1988). This has been regarded as a major mechanism of regulation of stomatal aperture by photosynthesis. However, Wong, Cowan & Farquhar (1979) noted a unique, strong positive correlation between gs and An. This asserted a hypothesis that the guard cells respond more directly to the mesophyll photosynthetic activity via some signals from the mesophyll (Wong et al. 1979; Messinger, Buckley & Mott 2006). Likewise, several empirical models of leaf photosynthesis assume close links between the An and gs (Ball, Woodrow & Berry 1987; Jarvis & Davies 1998; Yu et al. 2004).

In most amphistomatous leaves, stomata on the adaxial surfaces (upper surface) and those on the abaxial surfaces (lower surface) inhabit differing light environments in terms of both intensity and wavelength composition. The adaxial stomata are exposed to more direct radiation, whereas the abaxial ones are shaded by the leaf itself and receive the light transmitted through the mesophyll and reflected from the surroundings. Thus, we can expect differential responses to light stimuli between the adaxial and abaxial stomata in the same leaf.

Indeed, there have been several studies examining this possibility. Most of such investigations have been using in vitro experimental systems such as detached epidermal strips and guard cell protoplasts (Dale 1961; Pemadasa 1979, 1982; Travis & Mansfield 1981; Goh, Oku & Shimazaki 1995, 1997; Goh, Hedrich & Nam 2002). The results of these studies indicated that sensitivity to light is higher in the abaxial stomata than in the adaxial ones. However, because there remains a large gap between our knowledge obtained by in vitro studies and in vivo stomatal behaviors, in vivo studies are indispensable. Unfortunately, there are only several in vivo studies (Turner 1969; Turner & Singh 1984; Wong, Cowan & Farquhar 1985b; Yera et al. 1986; Lu, Quinones & Zeiger 1993). Yera et al. (1986), for example, measured the gs of intact leaves of Vicia faba with a diffusion porometer, and found that the maximum conductance of the adaxial surface was approximately two-thirds of that of the abaxial surface. In these in vivo studies, the ambient CO2 concentration (Ca) was kept constant. Thus, the Ci probably decreased with the increasing An of the leaves. Considering the convincing evidence that stomata respond to Ci and are insensitive to the CO2 concentration at the surface of the leaf (Scarth 1932; Sharkey & Raschke 1981a; Mott 1988), the control of Ci is essential in in vivo studies on stomatal light response.

Concerning the differences in light environment between the adaxial and abaxial stomata, it should be stressed here that the abaxial stomata always receive the light transmitted through the mesophyll (self-transmitted light), and that the self-transmitted light, rather than the light reflecting from the environment, is the major light source for the abaxial stomata. For example, under the growth conditions of Helianthus annuus in the present study, irradiance received by the abaxial side of the leaf was only about 10% of that of the self-transmitted light. Thus, it is highly probable that the abaxial stomata acclimate to the self-transmitted light, which is enriched in green light. However, the effects of the self-transmitted light have not yet been studied well. If we properly control the illumination conditions of the leaf, taking account of the effects of the self-transmitted light as well, differential responses between the adaxial and abaxial stomata will be more thoroughly described.

In this study, we set up a system to measure stomatal conductances and photosynthetic CO2 exchange rates of the respective surfaces of the leaf, separately and simultaneously. Using this system, we compared the light responses of the adaxial and abaxial stomata in intact sunflower leaves, keeping the Ci and other environment variables constant. We paid special attention to the effect of transmitted light. Four experiments were conducted. In experiment I, both sides of the leaf were illuminated simultaneously to quantify differences in the gs between the adaxial and abaxial stomata. In experiment II, we illuminated either side of the leaf, and found that the self-transmitted light induced stomatal opening more efficiently than the white light of the same PPFD directly applied to the stomata. We, therefore, further examined the effects of the transmitted light in experiment III. By placing one attached leaf between the white light source and the chamber, we obtained a ‘leaf-transmitted light’. Both the self-transmitted light and the leaf-transmitted light were more efficient in opening the abaxial stomata than the white light. As the transmitted light consists mostly of green light, we also evaluated the stomatal responses to monochromatic green light as well as to red and blue lights in experiment IV.

MATERIALS AND METHODS

Plant material and growth conditions

Seeds of sunflower (H. annuus cv. Russian Giant; Takii Seeds, Kyoto, Japan) were germinated in a plastic box for a week. Healthy seedlings were planted in vermiculite-containing pots (12 cm in diameter and 20 cm in height, one seedling per pot). The plants were grown in an environment-controlled chamber (Koitotron KG-50HLA-S; Koito Industries Ltd., Yokohama, Japan) at 60% relative humidity with a 14 h photoperiod at 25 °C and a 10 h dark period at 18 °C. Light was supplied by a bank of fluorescent lamps (FPR96EX-N/A Palook; National, Kadoma, Japan). The PPFD was measured with a photon sensor (LI-190; Li-Cor, Lincoln, NE, USA). The plants were supplied with approximately 350 µmol m−2 s−1 at canopy height. The plants for experiment IV were grown in another environment-controlled room at approximately 60% relative humidity with an 8 h photoperiod and a 16 h dark period at 23 °C. The PPFD at plant height was 250 µmol m−2 s−1.

The plants were watered daily and fertilized twice a week with 100 mL of Hoagland's solution containing 2 mm KNO3, 2 mm Ca(NO3)2, 0.75 mm MgSO4, 0.665 mm NaHPO4, 25 µm Fe-ethylenediaminetetraacetic acid, 5 µm ZnSO4, 0.5 µm CuSO4, 0.25 µm NaMoO4, 50 µm NaCl and 0.1 µm CoSO4. Fully expanded leaves were used in the experiments.

Gas exchange

A gas exchange system with a double-sided leaf chamber was constructed. In this system, N2 and O2 were mixed to the normal atmospheric concentrations (O2 : N2 = 1:4) using two mass flow controllers (MC-3000E; Lintec, Tokyo, Japan). Air humidity was controlled by bubbling air through water and passing through a copper-made condenser, whose temperature was controlled by circulating water. A mass flow controller (MC-3000E; Lintec) was used to add an appropriate amount of 5% CO2 from a gas cylinder to the gas stream. The resultant gas was divided into four pathways, each with a flow rate adjusted to 500 mL min−1. One pathway was used for the reference gas that passed through the reference cell of an infrared CO2 analyser (LI-7000, Li-Cor). The other three pathways were switched by solenoid valves and were alternately passed through the sample cell of the analyser. Two identical pathways were used to measure the transpiration and net photosynthetic rates for the adaxial and abaxial surfaces of the leaf separately, and the fourth one was the pathway for the reference gas to calibrate the reading of the sample gas. These four pathways had branches, and the cut ends of the branch tubes were placed under water. By adjusting the pressure inside the branches at 2 cm of water column, we balanced the pressure of all the pathways. The adjustment of the pressure in the branch tubes was made at 20 cm upstream of the chambers.

Two half-chambers, sandwiching the leaf, had windows of 6 cm2 (2 × 3 cm), allowing the independent control of the light environment of the two leaf surfaces separately. Temperature control of each of the half-chambers was achieved by circulating water in water jackets around the system. Leaf surface temperatures were measured by a pair of thermocouples attached to the adaxial and abaxial surfaces. In all experiments, no significant differences in the abaxial and adaxial surface temperatures were noted. Boundary layer conductances for water vapour at the flow rate of 0.5 L min−1, measured for wet filter paper, were 0.37 and 0.41 mol m−2 s−1 for the adaxial and abaxial half-chambers, respectively.

All data were logged on a personal computer. The CO2 and H2O concentrations and leaf temperature in each of the half-chambers were measured, and the rate of transpiration (E), Ci, An, gs and VPD were calculated for each of the leaf surfaces according to the method of von Caemmerer & Farquhar (1981). We were able to maintain the Ci and other conditions during the experiments by frequent adjustments of the CO2 injection rate and the temperature of the circulating water.

Prior to each series of the gas exchange measurements, leaves were kept in the dark for approximately 1 h. The PPFD was increased stepwise from the lowest level. At each PPFD level, data were taken after the gas exchange rates and the gs attained the steady state (in many cases, about 30 min after the shift of PPFD). The Ci was manipulated by varying the ambient CO2 concentration. The lag time of the system for stabilization after the change of ambient CO2 concentration was about 40 s. We were able to control the Ci at our expected level (for details, see Fig. A1 in the Appendix) via calculating the gas exchange parameters including the Ci by using the personal computer and adjusting the CO2 concentration every minute.

Light source

White light was provided by a fibre optic illuminator (COLD SPOT, PICL-NEX; NIPPON P-I CO. LTD, Tokyo, Japan) with a halogen projector lamp (15 V/150 W; Philips, Hamburg, Germany). The light was attenuated with a series of neutral density filters (Toshiba Corporation, Tokyo, Japan). In experiment III, we inserted an attached leaf between the fibre optic light emitter and the leaf chamber so that the leaf in the chamber receives a light transmitted through a leaf. No differences were noted in the transmittance of the leaf irrespective of whether the abaxial or adaxial surface was illuminated.

In experiment IV, blue monochromatic light was provided by four Luxeon Star LED lamps (Star/O, LXHL-NB98, 470 nm, maximum of 200 cd; Lumileds Lighting, LLC, San Jose, CA, USA), and red light was provided by nine super bright LED lamps (L-813SRC-E, 660 nm, maximum of 2800 mcd; Kingbright Electronic CO., LTD, Taiwan). The intensity was controlled by regulating the voltage from a direct current power supply (PMC 18-2, 0 V ∼ 18 V, 0 A ∼ 2 A; KIKUSUI Electronics Corp., Yokohama, Japan). Green monochromatic light was obtained by passing white light through a 540 nm interference filter (U-MNG2, BP530–550; Olympus, Tokyo, Japan). The leaf surface was evenly illuminated by the use of a diffuser. The spectra of the white light, the transmitted light and the three monochromatic lights were measured with a Li-1800 spectroradiometer (Li-Cor) using a laboratory-constructed integrating sphere.

Stomatal density and size

Microphotographs of the leaf surfaces were taken with a digital camera (VB-7010; KEYENCE, Osaka, Japan) attached to a microscope (OPTIPHOT-2; Nikon, Kawasaki, Japan). We selected 23 fully expanded leaves from 10 different plants, and for each side of these leaves, we took three different microphotographs at random. Samples were taken from the same area of the leaf laminas that were used for the gas exchange measurement. We counted the number of the stomata for each microphotograph (about 0.2255 mm2) to calculate the stomatal density. Because it was difficult to measure the short axes of stomata when the stomata were almost closed, we measured the long axes of stomata as the size of the stomata. We selected three different stomata at random, and measured the long axis of each stoma for each microphotograph.

Statistical analyses

Differences between means were analysed using Student's t-tests. Regression lines were obtained by the least squares method. Differences in the regression coefficient and in the intercept were detected using analysis of covariance (ancova). All statistical tests were carried out with the statistical software SPSS 12.0 J (SPSS, Inc., Chicago, IL, USA).

RESULTS

Stomatal density and size

The stomatal density on the AD was 77% of that on the AB, while the adaxial stomata lengths were approximately 98% of the abaxial values (Table 1).

Table 1.  Stomatal density and size
 Stomatal density (number of stomata mm−2)Long axis of stomatal aperture (µm)
  1. P-values are according to the two-tailed t-test between the adaxial and abaxial data.

Adaxial198 ± 68 (n = 69)26.7 ± 2.99 (n = 207)
Abaxial258 ± 70 (n = 69)27.3 ± 2.92 (n = 207)
P<0.0010.03
Ratio (adaxial/abaxial)0.770.98

Experiment I. Effects of simultaneous illumination from both sides

A bifurcated fibre optic illuminator was used in this experiment to illuminate both sides of the leaf with white light (spectrum, Fig. 1) at the same PPFD, varying from 30 to 1500 µmol m−2 s−1.

Figure 1.

Spectra of the leaf-transmitted light and the white light. A spectroradiometer Li-1800 (Li-Cor, Lincoln, NE, USA) was used to measure these spectra. The PPFDs are 20.18 and 22.28 µmol m−2 s−1 for the white light and transmitted light, respectively.

As the PPFD increased, the gs of both leaf surfaces increased. At any given PPFD, the gs of the AB (gs–AB) was always greater than that of the AD (gs–AD) (P < 0.05) (Fig. 2a). We calculated the gs* (Fig. 2b); gs*–AB was also much higher than gs*–AD (P < 0.05).

Figure 2.

Responses of the gs (a) and gs* (b) to white light simultaneously illuminated from both sides of the leaf. The PPFD was increased from the lowest value to the highest. Data are plotted against the PPFD illuminated to one leaf surface. Differences in the gs and gs* between the adaxial and abaxial surfaces were detected by Student's t-test. Data represent the means ± SD (leaf number ≧ 3). The Ci values are 301.5 ± 2.71 and 308.7 ± 6.32 µL L−1 for the adaxial and abaxial sides of the leaf. Leaf surface temperatures are 23.2 ± 1.48 and 23.1 ± 1.54 °C for the adaxial and abaxial surfaces, respectively. The VPD are 1.00 ± 0.19 and 0.85 ± 0.17 kPa in the upper and lower half-chambers. All these values are expressed as the means ± SD at the time points when the data of gs shown in the figure were taken (n = 29).

Experiment II. Effects of light from inside of the leaf

In this experiment, we initially illuminated the adaxial side of the leaf with white light at various PPFDs. After a dark period of approximately 1 h, we illuminated the abaxial leaf surface.

The transmittance of the leaf was measured with an integrating sphere and a quantum sensor. We assumed that the PPFD for the stomata on the other side of the illuminated surface is similar to that of the self-transmitted light, and plotted the data of gs against the self-transmitted PPFD. When the stomata received direct light, the data were plotted against the direct PPFD on the leaf surface (Fig. 3). Figure 3 insets show low PPFD data. Because the light intensities were different between the direct light and the self-transmitted light, and there were strong linear relationships between the gs and PPFD, we used ancova to compare the gs between light conditions. The slope was greater for the self-transmitted light than for the direct light in both gs–AD and gs–AB (P < 0.005), indicating that the stomata were more sensitive to the self-transmitted light. gs–AB was significantly higher than gs–AD at the same light intensity, confirming the results of experiment I. It should be noted, however, that the horizontal axis represents the PPFD of one side in Fig. 2 of experiment I. Because light was illuminated from both sides of the leaf in experiment I, the real PPFD received by the whole leaf was the sum of the PPFDs on both sides. Therefore, the An for the same PPFD value on the horizontal axis in experiment I was greater than that in experiment II (data not shown), which would explain the greater gs in Fig. 2 for the same PPFD on the horizontal axis (see experiment III).

Figure 3.

Relationships between the gs of the abaxial (upper panel) and adaxial (lower panel) surfaces to ‘real’ PPFD. The ‘real’ PPFD for the stomata denotes either the PPFD of the white light directly applied to the stomata, or the PPFD of the light transmitted by the leaf itself when the white light illuminated the other side of the leaf (self-transmitted light). The leaf was illuminated from only one side by the white light. inline image, ▪: gs in the white light directly applied to the stomata; inline image, □: gs in the self-transmitted light. The insets enlarge the low PPFD region. The regression lines are as follows: y = 0.0010x + 0.1176 (R2 = 1.000, P < 0.005) for gs–AB versus the self-transmitted light (upper panel), y = 0.00012x + 0.1138 (R2 = 0.9945, P < 0.005) for gs–AB versus the white light (upper panel), y = 0.00025x + 0.0274 (R2 = 0.9986, P < 0.005) for gs–AD versus the self-transmitted light (lower panel), and y = 0.000055x + 0.0215 (R2 = 0.9962, P < 0.05) for gs–AD versus the white light (lower panel). The slopes of the regression lines were compared by analysis of covariance (ancova). Data represent the mean ± SD (leaf number ≧ 4). The Ci are 300.2 ± 9.86 and 304.4 ± 8.49 µL L−1 for the adaxial and abaxial sides of the leaf, respectively. Leaf surface temperatures are 22.4 ± 1.03 and 22.4 ± 1.11 °C for the adaxial and abaxial surfaces, respectively. The VPD are 0.99 ± 0.14 and 0.84 ± 0.13 kPa in the upper and lower half-chambers. All these values are expressed as the means ± SD at the time points when the gs data shown in the figure were taken (n ≧ 32).

Experiment III. Effects of the light transmitted through the leaf

In order to confirm whether the stomata are more sensitive to the transmitted light, we placed an attached leaf in the light path so that the light received by the leaf in the chamber was the light transmitted through the leaf (the light thus obtained is called ‘leaf-transmitted light’ in order to distinguish the light from ‘self-transmitted light’). The spectra of the white light and the transmitted light are compared in Fig. 1. After the gas exchange measurements with the leaf-transmitted light, the leaf was kept in the dark in the chamber for approximately 1 h, and the same surface was illuminated with white light. We increased the PPFD so that the stomata on the opposite side of the leaf receive the leaf-transmitted light at PPFD levels comparable with those received by the stomata on the illuminated sides in the preceding two series of measurements. The same procedure was repeated for the other side of the leaf. In this experiment, therefore, the stomatal conductances of both sides were measured with the white light or the leaf-transmitted light illuminated from the same side or the opposite side. Results are presented in Fig. 4.

Figure 4.

Relationships between the stomatal conductances of the adaxial or abaxial surfaces to the ‘real’ PPFD of the self-transmitted light, leaf-transmitted light or white light. For the definitions of the self-transmitted light and white light, see the legend for Fig. 3. The leaf-transmitted light denotes the light transmitted by the other attached leaf located between the white light source and the chamber. □, inline image, ▪: gs of the abaxial side obtained with the leaf-transmitted light, self-transmitted light and white light, respectively; ▵, inline image, ▴: gs of the adaxial side obtained with the leaf-transmitted light, self-transmitted light and white light, respectively. Data represent the means ± SD (leaf number ≧ 3). The regression lines are as follows: y = 0.00084x + 0.0887 (R2 = 0.9858, P < 0.001) for gs–AB versus the self-transmitted light, y = 0.00041x + 0.0894 (R2 = 0.9869, P = 0.073) for gs–AB versus the leaf-transmitted light, y = 0.00018x + 0.0916 (R2 = 0.9868, P = 0.073) for gs–AB and white light, y = 0.00024x + 0.0182 (R2 = 0.9646, P < 0.001) for gs–AD versus the self-transmitted light, y = 0.00018x + 0.0165 (R2 = 0.9142, P = 0.189) for gs–AD versus the leaf-transmitted light, and y = 0.00016x + 0.0132 (R2 = 0.9646, P = 0.269) for gs–AD versus the white light. The slopes and intercepts of the regression lines were compared by analysis of covariance (ancova). The Ci are 306.8 ± 6.48 and 304.6 ± 2.97 µL L−1 for the adaxial and abaxial sides of the leaf. Leaf surface temperatures are 22.6 ± 0.98 and 22.6 ± 1.05 °C for the adaxial and abaxial surfaces, respectively. The VPD are 0.98 ± 0.18 and 0.91 ± 0.12 kPa in the upper and lower half-chambers. All these values are expressed as the means ± SD at the time points when the gs data shown in the figure were taken (n ≧ 46).

We used ancova to compare the gs under these three light conditions for each side of the leaf. For gs–AB, the slope for the self-transmitted light was greatest, followed by those for the leaf-transmitted light and for the white light, in this order. In other words, the abaxial stomata were most sensitive to the self-transmitted light, and least sensitive to the white light (Fig. 4). The regression coefficient for the relationship between gs–AD and the PPFD of the leaf-transmitted light, or that for the relationship between gs–AD and the white light was not statistically significant (P > 0.1), so that a comparison among the three light conditions was not available (Fig. 4). However, the sensitivity of the adaxial stomata to both of the transmitted light was low.

Whole leaf An was calculated as the sum of the photosynthetic rates measured for the adaxial and abaxial sides (Fig. 5). When all the data were pooled, the correlation between the An and gs–AB was stronger than that between the An and gs–AD. In Fig. 5, regardless of the lighting conditions, a positive relationship was noted between gs–AB and An, suggesting that the light response of the abaxial stomata involved a photosynthesis-dependent effect. Although the slopes of the regression lines for the leaf-transmitted light and of the self-transmitted light did not differ significantly (P = 0.238), the intercepts of the lines differed significantly (P < 0.005, Fig. 5). Thus, when compared at the same An, gs–AB by the leaf-transmitted light tended to be greater than those by the self-transmitted light (as the linear regression of gs–AB on An in the white light was not significant, the comparison here did not include the white light data). On the other hand, the relationships between the An and gs–AD were weak. For the same An, gs–AD by the leaf-transmitted light tended to be greater than gs–AD by the self-transmitted light.

Figure 5.

Relationships between the whole leaf An and the abaxial gs (upper panel) or adaxial gs (lower panel). An denotes the sum of the photosynthetic rates of both sides. The regression lines are y = 0.0043x + 0.0776 (R2 = 0.6981, P < 0.001) for gs–AB versus An in the self-transmitted light (open squares), y = 0.0117x + 0.0981 (R2 = 0.4805, P < 0.05) for gs–AB versus An in the leaf-transmitted light (grey squares), and y = 0.0045x + 0.0853 (R2 = 0.3274, P = 0.107) for gs–AB versus An in the white light (solid squares). The slopes and intercepts of the regression lines were compared by analysis of covariance (ancova). Neither of the regression coefficients of the regression lines of gs–AD versus An was significant.

Experiment IV. Effects of monochromatic light

Because transmitted light is enriched in green light, we further investigated the green light response of the stomata. Leaves were irradiated directly with monochromatic light. The spectra of the monochromatic lights are shown in Fig. 6a. For gs–AB, the slope for the blue light was greatest (P < 0.001) (Fig. 6b), while those for the green light and red light were not significantly different (P > 0.1). The sensitivity to monochromatic light of the adaxial stomata was generally low (Fig. 6c). The slope of the regression line for the blue light was, however, significantly higher than that for the red light (P < 0.05). Interestingly, the adaxial stomata appeared to be insensitive to the green light. We also showed the relationship between the gs and the whole leaf An for each monochromatic light in Fig. 6d, e. A positive relationship was also noted between the gs and An, regardless of the lighting conditions (except for the green light on the adaxial side), again suggesting the involvement of the photosynthesis-dependent effect on stomatal opening. When compared at the same An, gs–AB by the blue light was greatest, followed by gs–AB by the green light and the red light, in this order (P < 0.01). Analogically, for the same An, gs–AD by the blue light was greater than gs–AD by the red light. The relationship of the An and gs–AD by the green light was weak.

Figure 6.

Responses of gs to monochromatic light. The monochromatic light was irradiated directly to each side of the leaf. (a) Spectra of the blue, green and red monochromatic lights; stomatal conductances of the abaxial (b) and adaxial (c) surfaces plotted against the PPFD of the blue light (blue triangle), green light (green square) or red light (red circle), respectively; (d, e) gs plotted against the whole leaf An driven by the monochromatic light. Data represent the means ± SD (leaf number ≧ 3). Data that represent the regression lines in (b) are y = 0.0023x + 0.0732 (R2 = 0.9337, P < 0.005), y = 0.0007x + 0.0883 (R2 = 0.8462, P < 0.01), and y = 0.0007x + 0.0837 (R2 = 0.8993, P < 0.005) for gs–AB versus the blue light, green light and red light, respectively. The regression lines in (c) are y = 0.00008x + 0.0046 (R2 = 0.9373, P < 0.005) and y = 0.00005x + 0.0041 (R2 = 0.9700, P < 0.001) for gs–AD versus the blue light and red light, respectively. The regression for gs–AD versus the green light was not significant (P > 0.1). The symbols are same with (b) and (c). The regression lines in (d) are y = 0.0331x + 0.1528 (R2 = 0.9657, P < 0.005), y = 0.0281x + 0.1064 (R2 = 0.8983, P < 0.05), and y = 0.0184x + 0.0999 (R2 = 0.9766, P < 0.001) for gs–AB versus the blue light, green light and red light, respectively. The regression lines in (e) are y = 0.0028x + 0.0074 (R2 = 0.9892, P < 0.001) and y = 0.0015x + 0.0039 (R2 = 0.9139, P < 0.05) for gs–AD versus the blue light and red light, respectively. The regression for gs–AD versus the green light was not significant (P > 0.1). The slopes and intercepts of the regression lines were compared by analysis of covariance (ancova). The Ci are 305.7 ± 6.35, 302.5 ± 3.92 and 300.8 ± 8.38 µL L−1 for the red light, green light and blue light from the adaxial side, and 299.2 ± 6.34, 304.1 ± 6.58 and 303.5 ± 5.28 µL L−1 for the red light, green light and blue light from the abaxial side, respectively. Leaf surface temperatures are 22.6 ± 0.38, 22.0 ± 0.26 and 22.1 ± 0.32 °C for each monochromatic light measurement of the adaxial surface, and 22.4 ± 0.39, 22.6 ± 0.25 and 22.5 ± 0.53 °C for each monochromatic light measurement of the abaxial surface. The VPD are 0.70 ± 0.02, 0.69 ± 0.11 and 0.70 ± 0.15 kPa in the upper half-chamber, and 0.74 ± 0.15, 0.77 ± 0.12 and 0.78 ± 0.13 kPa in the lower half-chambers for the red, green and blue monochromatic lights. All these values are expressed as the means ± SD at the time points when the gs and An data shown in the figure were taken (n ≧ 24).

DISCUSSION

The present results showed that the abaxial stomata were more sensitive to the white light than the adaxial stomata even after correcting the difference in the stomatal density (Fig. 2). Higher sensitivities of the abaxial stomata were also observed for the self-transmitted light (Figs 3 & 4), the leaf-transmitted light (Fig. 4), and the monochromatic blue, red and green lights (Fig. 6). Howard & Donovan (2007) found that sunflower abaxial stomata do not close completely even in the dark. Although we did not routinely measure the gs in the dark, the results of experiment IV (Fig. 6) indicated that the larger gs in the abaxial stomata at the lowest PPFDs would be partly due to dark opening. From the differences in the slopes in Figs 2–4 and 6 between the abaxial and adaxial stomata, however, higher sensitivities of the abaxial stomata were clear.

Some in vivo studies with sunflower (H. annuus L. cv. Hysun 30), sorghum (Sorghum bicolor (L.) Moench cv. RS610) and tobacco (Nicotiana tabacum L.) also showed similar differences in the sensitivity between the adaxial and abaxial stomata (Turner 1969; Turner & Singh 1984), but the differences were not clear in broad bean (V. faba L.) (Yera et al. 1986) or pima cotton (Gossypium barbadense L.) (Lu et al. 1993). In vitro studies using epidermal peels of broad bean (Goh et al. 2002), Commelina communis L. (Travis & Mansfield 1981; Pemadasa 1982) and cotton (Dale 1961), or guard cell protoplasts of broad bean (Goh et al. 1995, 1997), also showed a higher light sensitivity in the abaxial stomata than in the adaxial ones. The aims of these studies were not necessarily to clarify the mechanisms responsible for the different light sensitivity of the abaxial and abaxial stomata. Although Goh et al. (1995) postulated that the blue-light photoreceptor exists at a higher concentration in the abaxial guard cells than in the adaxial guard cells, this possibility has not been tested and the mechanisms are largely unknown. We discuss the present results in the light of the mechanisms for the differential light sensitivity of the adaxial and abaxial stomata. We highlight the effects of photosynthesis and transmitted light on stomatal behaviour.

Among the white light, the leaf-transmitted light and the self-transmitted light, supplied at the same PPFD incident on the abaxial stomata, the self-transmitted light resulted in the highest gs, followed by the leaf-transmitted light and the white light, in this order (Fig. 4). The An was highest for the self-transmitted light, because the light given to the other side was much stronger. Thus, the results suggest the effect of mesophyll photosynthesis on stomatal opening. At a given PPFD on the horizontal axes of Figs 2 and 3, the higher gs in Fig. 2 than that in Fig. 3 can also be explained by the differences in the rate of mesophyll photosynthesis.

As we kept the Ci nearly constant during all the experiments, the effect of photosynthesis was not mediated by the changes in the Ci. Demonstrating a linear relationship between the An and the gs, Wong, Cowan & Farquhar (1985a,c) and Wong et al. (1979, 1985b) suggested that photosynthesis could affect stomatal opening directly. They hypothesized that stomata respond to an unknown photosynthetic metabolite in the leaf mesophyll tissue (Wong et al. 1979). Conversely, it has been shown that the linear relationships between the An and gs do not hold in rubisco small subunit (rbcS) antisense plants (Stitt et al. 1991; Hudson et al. 1992; von Caemmerer et al. 1997, 2004), suggesting that regulatory mechanisms may be more complex than initially envisaged.

The effects of mesophyll photosynthesis were more prominent in the abaxial stomata than in the adaxial ones (Figs 5 & 6d, e). As the structures of palisade tissue and spongy tissues are different, the photosynthate translocation and/or the signal transduction pathways for the regulation of the stomatal movements may be different between the adaxial and abaxial stomata. Based on their study of the differential responses of the abaxial and adaxial guard cells to abscisic acid and Ca2+, Wang, Wu & Assmann (1998) also made a similar argument. The difference in the sensitivity to mesophyll photosynthesis between the abaxial and adaxial stomata may, at least, partly contribute to the difference in the light sensitivity between the adaxial and abaxial stomata. The adaxial and abaxial stomata would be utilized as an experimental system useful for the identification of the mesophyll to stomata signal(s).

For the abaxial stomata, the leaf-transmitted light, which was the light transmitted through the other leaf and thus contained much green light (Fig. 1), induced stomatal opening more efficiently than the white light (Fig. 4). The abaxial stomata certainly opened when the monochromatic green light was applied (Fig. 6b). However, the monochromatic blue light was more efficient, and the white light contained more blue light than the leaf-transmitted light (Fig. 1). Therefore, the present results of monochromatic lights cannot explain the effects of the leaf-transmitted light. Some in vitro studies showed that the mixing ratio of two (or more) monochromatic lights is an important factor affecting the stomatal opening. For example, the extent of blue-light-stimulated stomatal opening depends on the level of the red light background (Kinoshita et al. 2001; Talbott et al. 2003a,b). Moreover, recent studies with V. faba (Frechilla et al. 2000) and Arabidopsis thaliana (Zeiger et al. 2002; Talbott et al. 2006) have shown that stomatal opening induced by blue light can be reversed by green light. Under a continuous illumination, the reversing effect was proportional to the fluence rate of green light, with a full reversal observed when the green and blue lights were applied at a 2:1 ratio (Zeiger et al. 2002; Talbott et al. 2006). The effect would not explain the present data. In the present study, the ratio of blue light/green light/red light was clearly different between the white light and the transmitted light. The ratio of these lights could be an important determination of the stomatal opening. Interestingly, the adaxial stomata seemed insensitive to the monochromatic green light (Fig. 6c), which might, at least, partly answer the question why the adaxial stomata showed little response to the leaf-transmitted light. Light environment during leaf growth may be responsible for the scarcity of the response of the adaxial stomata to green light. To prove this possibility, we are currently conducting growth experiments with manipulated light environments during leaf growth.

In this study, we clarified that the sensitivity to light was higher in the abaxial stomata than in the adaxial stomata, and that the abaxial stomata were more sensitive to the transmitted light than to the white light. It is noteworthy that, because the intensity of the blue light in the transmitted light was lower than that in the white light, the stomata are more sensitive to the light composed of some monochromatic lights at a certain mixing ratio than to a particular monochromatic light. Moreover, our results indicated that a photosynthesis-dependent effect is involved in stomatal light response. Because we maintained the Ci during all experiments, these results support the hypothesis by Wong et al. (1979) that the stomata respond to an unknown photosynthetic metabolite in the mesophyll. As the mechanisms for the different sensitivity of the adaxial and abaxial stomata, we were able to identify two facts: (1) the adaxial stomata were almost insensitive to the monochromatic green light, and (2) the effects of mesophyll photosynthesis were small in the adaxial stomata. Such qualitative differences in the light responses between the adaxial and abaxial stomata were reported for the first time. Comparing the different responses of the adaxial and abaxial stomata can be a useful point of view with which we may investigate stomatal behaviours further.

ACKNOWLEDGMENTS

We thank Dr. Kouki Hikosaka for the loan of a spectroradiometer, and Louis Irving for editing the text. This study was supported by Japan Society for the Promotion of Science (No. 16207002).

Appendix

image

Figure A1.  Time-dependent changes in gas exchange parameters for the two leaf surfaces of the Helianthus annuus leaf kept in the dark and then illuminated from either the adaxial (a–c) or the abaxial (d–f) side. (a, d) ▪ and □ represent the CO2 concentration in the chamber (Ca) in the upper and lower chambers, and inline image and ○ represent the Ci of the adaxial and abaxial sides, respectively; (b, e) inline image and ○ represent the An of the adaxial and abaxial sides; (c, f) inline image and ○ represent the gs of the adaxial and abaxial sides, respectively. Two separate leaves were used. One was illuminated from the adaxial side while the other was illuminated from the abaxial side. The leaf in the chamber was kept in the dark in chambers containing CO2 at about 320 µL L−1 for at least 30 min, and then illuminated with white light at 250 µmol m−2 s−1. The CO2 concentration of the inlet air was manipulated every minute to keep the Ci as close as 300 µL L−1, and the leaf temperature was kept at 21 °C. The VPD was kept at around 0.74 kPa. The steady-state level of gs was attained by 40 min. In the experiments shown in the main part, the PPFD increased stepwise; once illuminated, the gs tend to attain the steady-state level much faster. We manipulated Ca in the inlet air to keep the Ci as constant as possible except during the dark pre-treatment.

Ancillary