Leaf surface wetness that occurs frequently in natural environments has a significant impact on leaf photosynthesis. However, the physiological mechanisms for the photosynthetic responses to wetness are not well understood. The responses of leaf CO2 assimilation rate (A) to 72 h of artificial mist of a wettable (bean; Phaseolus vulgaris) and a non-wettable species (pea; Pisum sativum) were compared. Stomatal and non-stomatal limitations to A were investigated. A 28% inhibition of A was observed in the bean leaves as a result of a 16% decrease in stomatal conductance and a 55% reduction in the amount of Rubisco. The decrease of Rubisco was mainly due to its partial degradation. In contrast to the bean leaves, a 22% stimulation of A was obtained in the 72 h mist-treated pea leaves. Mist treatment increased stomatal conductance by 12.5% and had no effect on the amount of Rubisco. These results indicated that a positive photosynthetic response to wetness occurred only in non-wettable species and is due to the change in stomatal regulation.
In a natural environment, the leaf surfaces of plants frequently are wetted by rain, fog and dew. Surface wetness induces different changes in leaf photosynthesis among species, because leaf surface wettability varied greatly, that is, from being covered almost completely by water to being water repellent (Smith & McClean 1989; Brewer, Smith & Vogelmann 1991; Brewer & Smith 1997). In alpine and sub-alpine plants, natural dew depressed A by 77% in species having wettable leaves, whereas A was stimulated by 14% in species having non-wettable leaves (Smith & McClean 1989).
The different responses of A to wetting have been attributed to differences in the temporal impact of surface water on the stomata. Water droplets on the leaves may reduce the water vapour deficit (Smith & McClean 1989). Wetness may thus stimulate stomatal conductance and thus A in ‘non-wettable’ species. The reduction of A by wetness in the ‘wettable’ species may be partly due to the physical blockage of stomata by the surface water (Brewer & Smith 1995). In these cases, photosynthetic responses will be reversible, i.e. A will fully recover when water droplets are removed from the leaf surface.
However, the reduction of A by wetness is irreversible when leaves of a ‘wettable’ species (bean; Phaseolus vulgaris) are subjected to long-term mist exposure. A reduction in A was observed in 24 h mist-exposed bean leaves after removing water droplets (Ishibashi & Terashima 1995; Ishibashi, Usuda, & Terashima 1996). However, little is known about the mechanism of the irreversible response or whether it differs among species.
The ‘irreversible’ response of A to surface wetness may be caused by some alterations in the two main limiting processes: (1) CO2 diffusion processes; and (2) CO2 fixation processes mediated by Rubisco. Surface wetness should affect the signals controlling stomatal regulation, such as water loss rate (Mott & Parkhurst 1991; Monteith 1995) and leaf water potential (Comstock & Mencuccini 1998). Surface wetness may affect leaf internal CO2 diffusion through the alteration of leaf morphology. A morphological change by wetness (Kimura 1987) may involve an alteration in mesophyll anatomy such as chloroplast distribution and cell wall thickness, which determines internal conductance (Evans et al. 1994; Kogami et al. 2001).
The objective of this study is to clarify the mechanisms for the response of A to surface wetness for species with different wettability. We focused on the ‘irreversible’ effects of wetting on photosynthetic machinery. A 72 h mist pretreatment was conducted to compare the A between bean (Phaseolus vulgaris) and pea (Pisum sativum). Bean leaves have a large and flat surface that is wetted with ease (wettable), whereas pea leaves have a smooth and waxy surface that is difficult to wet (non-wettable, Tukey 1970). Gas exchange parameters and the amount of Rubisco in the leaves were analysed as possible factors underlying the photosynthetic response. We obtained different responses of A between bean and pea, which was governed by changes in stomatal limitation and the amount of Rubisco.
MATERIALS AND METHODS
Seeds of bean (Phaseolus vulgaris L. cv. Kurosando-saito) and pea (Pisum sativum L. cv. Denko) were obtained commercially (Takii Seeds, Kyoto, Japan). Each of the seedlings was transplanted to vermiculite in a 600-cm3 pot and grown in a growth chamber (LPH-1000-NCR; Nippon Medical and Chemical Instruments, Osaka, Japan). The chamber was kept at 25 °C, 80% relative humidity, 300 µmol m−2 s−1 PPFD, and 12 h photoperiod. The light was provided by four high-intensity discharge lamps (DR400/T-L; Toshiba, Tokyo, Japan). Plants were fertilized daily with a half-strength Hoagland solution (Hewitt & Smith 1975).
For the wetness pretreatment, the mist of deionized water (pH = 5.1) was applied uniformly to the plants from four spray nozzles at a rate of 470 cm3 h−1. The soil surface of the pots was covered with a plastic film to prevent the soil from being wetted by the mist. At the start of the mist pretreatment the leaf age was 10 d for the primary leaves of bean and 7 d for the pea leaves.
Leaf surface properties
The apparent contact angle of water droplet (θ) was measured with a contact angle-meter (model G-1; Erma Inc., Tokyo, Japan) for a 5-mm3 water droplet placed onto the leaf surface. A lower value of θ indicates a less spherical water droplet and thus a more-wettable surface (Brewer et al. 1991). To measure the amount of water on the leaves, 50 cm3 of water was sprayed on a pre-weighed leaf with a cone nebulizer, and then the leaf was immediately re-weighed.
Water absorption during the mist pretreatment may affect leaf water potential and thus stomatal regulation. Therefore, we checked whether bean and pea leaves were able to absorb water directly from the leaf surface, using fluorescence microscopy. The upper surface of the intact leaves was sprayed with a 0.1% (w/v) aqueous solution of fluorescent brightener-28 (Sigma-Aldrich, St Louis, MO, USA), which binds to the polysaccharides of the cell walls and emits blue fluorescence under a wavelength of 350 nm (Munné-Bosch, Nogués & Alegre et al. 1999). After 30 min, these leaves were washed in distilled water, and then their transverse sections were stained in distilled water in order to observe them under a fluorescence microscope (BX51; Olympus, Tokyo, Japan) at 350 nm.
Gas exchange measurements
Gas exchange measurements were made with a system described by Hanba, Kogami & Terashima (2002) for intact leaves. Measurements were repeated twice after a 10-min acclimation to 740 µmol m−2 s−1 PPFD for each of the leaves. The light source was a quartz-halogen lamp (350 W) with a reflector for far-red light. In the course of the measurements, leaf temperature and ambient CO2 partial pressure was kept at 24.0 ± 1.0 °C and 35 ± 0.4 Pa, respectively.
Before the mist pretreatment, the temporal effects of surface water on CO2 assimilation rate (A) were detected by measuring the time-course of A after spraying distilled water on the upper surface of the leaves. In the bean leaves, water droplets covered their surfaces almost completely. The water droplets on the leaves were lost by evaporation after 10 min. For the pea leaves, the water droplets covered only a small part of the leaves.
For the measurements of the leaves from the 24–72 h mist pretreatment, water droplets on the leaves were carefully removed using filter papers, and then the gas-exchange was measured after 10-min air-drying in laboratory conditions. The stomatal response to vapour pressure deficit (VPD) was analysed by changing the vapour pressure of the air entering the assimilation chamber.
Detection of Rubisco and its degradation products
Three leaf discs (1.8 cm2 × 3) were obtained from the central part of the lamina, and immediately homogenized at 4 °C in a 50-mm HEPES buffer (pH 7.5) containing 0.2% triton X-100, 0.7% 2-mercaptoethanol, 2 mm monoiodoacetic acid, 25% glycerol, 6% lithium dodecyl sulfate (LDS), and 1% polyvinyl polypyrrolidone (PVP). No protease inhibitor was added to the buffer. These homogenates were immediately centrifuged, and their supernatants (0.2 cm3) were boiled for 5 min with the Laemmli buffer (Laemmli 1970). The proteins in the supernatants were separated using 14% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The SDS-PAGE gels were stained with a Coomassie Brilliant Blue G-250 (Bio-safe Coomassie; Bio-Rad, Hercules, CA, USA). The amount of Rubisco subunits was determined spectrophotometrically by scanning the gel with a scanner (GT7600U; Epson, Tokyo, Japan), analysed using software (Densitograph; ATTO, Tokyo, Japan).
After the electrophoretic separation by SDS-PAGE was completed, the proteins were transferred onto polyvinylidene fluoride membranes (Immune-blot PVDF membrane; Bio-Rad) using a semi-dry system (NA-1515; Nippon-Eido, Tokyo, Japan). Rubisco was detected with an antibody against Rubisco (1/3000 dilution) and visualized using an alkaline phosphatase kit (AP conjugate substrate kit; Bio-Rad). Band analysis was performed as described in the previous section.
Leaf morphology and biochemistry
Stomatal density and size of stomata were measured to obtain the effect of the mist pretreatment on stomata of bean and pea. A mixture of a silicone rubber monomer (KE-10; Shinetsu Chemical, Tokyo, Japan) and a catalyst (CAT-RA; Shinetsu Chemical) was poured on the upper and lower side of the leaves (n = 5). The impression was cleaned using 99% ethanol, and then a secondary replica was taken by applying nail varnish. Ten stomata were randomly taken to determine the size of stomata for each of the replicas, analysed with a light micrograph using software (NIH Image; National Institute of Health, Bethesda, MD, USA).
Chlorophyll was extracted in 4 cm3 dimethylformamide from a leaf disc (1.8 cm2), obtained from the central part of the lamina. Chlorophyll concentration was determined spectrophotometrically from the absorbance at wavelengths of 663.8 and 646.8 nm (U-1800; Hitachi, Tokyo, Japan) following Porra, Thompson & Kriedemann (1989).
For the analysis of soluble protein, a leaf disc (1.8 cm2) was obtained from the central part of the lamina, and immediately homogenized at 4 °C in a 50-mm HEPES buffer (pH 7.5) containing 0.7% polyethylene glycol 2000 (w/v), 0.2% PVP, and 1 mm phenylmethylsulfonyl fluoride (PMSF). These homogenates were immediately centrifuged, and the amount of protein was determined for their supernatants using the Bradford method.
Leaf discs (1.8 cm2) were cut from the central part of the lamina, kept in the dark for 15 min to measure initial chlorophyll fluorescence (Fv/Fm) using a portable fluorometer (Plant Efficiency Analyser; Hansatech, Kings Lynn, Norfolk, UK).
We checked the effect of the mist pretreatment on mesophyll anatomy and internal conductance for CO2 diffusion (gi) in the bean leaves. These effects were expected only for the ‘wettable’ species, because a previous study for barley cultivars showed a significant alteration in leaf morphology in the ‘wettable’ mutant but much less alteration in the ‘non-wettable’ wild type (Takehana et al. unpublished work).
Leaf mesophyll anatomy was analysed using light microscopy, as described in Hanba et al. (2002). Transverse slices 1 mm × 3 mm were cut from the central part of the lamina, fixed in 2.5% glutaraldehyde and 2% osmium tetroxide, then embedded in Spurr's resin (Spurr 1969). Transverse sections, 700 nm thick, were stained in a 1% toluidine blue solution, and light micrographs were taken with a digital camera (DP11; Olympus, Tokyo, Japan). Anatomical characteristics were determined from these 150 µm-wide, digitized images using software (NIH Image).
Internal conductance to CO2 diffusion was estimated by concurrent measurements of gas exchange and carbon isotope ratio as described by Hanba et al. (2002). The conditions for the gas exchange measurements were as described in the previous section. Carbon isotope ratios were measured with a mass spectrometer (Finnigan MAT 252, Bremen, Germany).
Effects of treatment were tested with a t-test using StatView (SAS Institute, Cary, NC, USA) at a probability level of P < 0.05.
Leaf surface properties related to the surface wettability were characterized for bean and pea (Table 1). The bean leaves had a much lower apparent contact angle of a water droplet (θ) compared with those of the pea (Table 1). In addition, the bean leaves retained 1.7-fold more water than pea leaves (Table 1). These measurements showed that the bean leaves had more wettable surfaces in comparison with the pea leaves.
Table 1. Leaf surface properties of the upper and lower sides of the leaves of bean (Phaseolus vulgaris) and pea (Pisum sativum). Leaf age was 10 d in bean and 7 d in pea. A high contact angle of water droplets indicates a more water repellent surface
Data are mean (SD) for n = 5.
Leaf area (cm2)
Amount of retained water (mg cm−2)
Contact angle of water droplet (°)
The bean leaves that were sprayed with an aqueous solution of fluorescence showed blue fluorescence on epidermal and mesophyll cell walls, whereas no blue fluorescence was observed in the pea leaves (Fig. 1). This result indicates that surface water entered the mesophyll cell through the epidermis in the bean leaves but not in the pea leaves.
The temporal effects of surface water on leaf CO2 assimilation rate (A) was compared between bean and pea (Fig. 2). A remarkable reduction of A was obtained in the wet leaves of bean, and then A recovered rapidly with the evaporation of the surface water. Although the surface water on the bean leaves was lost almost completely after 10 min, the reduction of A was still observed. In contrast, no reduction of A was obtained in the wet leaves of pea.
The effect of 72 h mist pretreatment on leaf CO2 assimilation rate (A) and stomatal conductance (gs) was compared between bean and pea (Fig. 3). In control leaves of bean and pea, both A and gs were constant during the experiment (data not shown). A significant inhibition of A was obtained in bean, contrasting with the stimulation of A in pea (Fig. 3a). A 72 h pretreatment inhibited A by 28% in bean, whereas it enhanced A by 22% in pea. The response of gs in bean was different from that in pea (Fig. 3b). The 72 h pretreatment decreased gs by 16% in bean, whereas gs was stimulated by 12.5% in pea (Fig. 3b).
The responses of stomatal conductance (gs) to VPD were compared between no-mist and 72 h mist-treated leaves to investigate the effect of mist pretreatment to stomatal regulation (Fig. 4). The mist-treated bean leaves showed lower gs in comparison with the control leaves from 500 to 1700 Pa VPD. In contrast, the mist-treated pea leaves had higher gs than the control leaves. The sensitivity of stomata to VPD was similar between species in the control leaves, whereas in the mist-treated leaves, pea leaves showed high sensitivity to VPD in comparison with the bean leaves.
The amount of proteins of the leaf extracts was detected by SDS-PAGE using a Coomassie-blue stain, and the degradation products of Rubisco in the leaf extract was detected using a western blot (Fig. 5). In the Coomassie stain, a most abundant protein of 53 kDa was attributed to the Rubisco large subunit (LSU). A 15-kDa band was attributed to the Rubisco small subunit (SSU). Mist pretreatment reduced the amount of LSU and SSU in bean, whereas no such reductions were observed in pea. Western blot data showed that with the major band of 53 kDa, fragments of 42, 31 and 23 kDa were recognized in the extracts of both of the species. They were likely to be the degradation products of Rubisco. Mist pretreatment increased the intensities of these fragments in the bean leaves, whereas the pretreatment effect was less in the pea plants.
The effect of mist pretreatment on the amount of Rubisco was quantified using a Coomassie stain of the SDS-PAGE gel (Fig. 6). At the start of the pretreatment, the amount of LSU was 3.3 ± 0.2 and 4.2 ± 0.3 g m−2 in the bean and the pea leaves, respectively (mean ± SD, n = 5). The amount of LSU was reduced by 55% after the 72 h mist pretreatment in the bean leaves, but no reduction was obtained in the pea leaves. Similarly, the amount of SSU was decreased by 34% in the 72 h mist-treated bean leaves, whereas no change was obtained for the pea leaves.
Rubisco degradation was analysed using Western blot (Fig. 7). The breakdown products of 42 and 23 kDa were increased after 72 h mist pretreatment in bean, whereas no effect of pretreatment was observed in pea. The amount of these breakdown products was higher in bean than those in pea.
The effect of 72 h mist pretreatment on stomatal distribution was examined for bean and pea. Leaf morphology, chlorophyll concentration and Fv/Fm were little affected for both species (Table 2). On the other hand, stomatal density was increased in pea but not in bean. Increases in the size of stomata and in the amount of soluble protein were obtained in the bean only.
Table 2. Effects of 72 h mist pretreatment on leaf characteristics of bean (Phaseolus vulgaris) and pea (Pisum sativum)
72 h mist
72 h mist
Data are mean (SD) for n = 5. Means followed by different letters are statistically different, tested by t-test (P < 0.05).
Upper surface (no. mm−2)
Lower surface (no. mm−2)
96 (10) a
114 (8) b
Upper surface (µm)
25.6 (2.0) a
32.5 (2.1) b
Lower surface (µm)
24.5 (2.0) a
28.8 (1.8) b
Leaf dry mass per area (g m−2)
Chlorophyll a+b (µmol m−2)
Soluble protein (g m−2)
3.10 (0.18) a
3.92 (0.70) b
Leaf internal conductance (gi) and related mesophyll anatomy were estimated in the bean plants only to examine the mesophyll limitation to A(Table 3). The value of gi was not affected by the mist pretreatment in the bean leaves. Mesophyll thickness and surface area of mesophyll cells exposed to intercellular airspaces (Smes) were unaffected by the pretreatment, whereas the surface area of chloroplasts facing to intercellular airspaces (Sc) declined slightly.
Table 3. The effects of 72 h mist pretreatment on the leaf mesophyll anatomy and internal CO2 conductance in bean (Phaseolus vulgaris). Smes or Sc indicates surface area of mesophyll or chloroplasts exposed to intercellular airspaces. No analysis was done for the pea leaves, because mist treatment causes no morphological alteration in pea (data not shown)
72 h mist
Data indicate mean ± SD (n = 5). Difference of means was tested as described in Table 1.
Mesophyll thickness (µm)
Mesophyll porosity (m3 m−3)
0.38 (0.04) a
0.45 (0.02) b
14 (1) a
11 (1) b
Internal conductance (mol m−2 s−1)
Our finding that exposing bean leaves to 5 min of mist resulted in a significant and temporal reduction of the CO2 assimilation rate (A, Fig. 2) suggest that stomatal blockage by the film of surface water strongly inhibits CO2 diffusion in ‘wettable’ species. This result supports previous reports that, at least for ‘wettable’ species, a temporal response of stomata to water droplets on the leaves had a significant impact on photosynthesis (Smith & McClean 1989; Brewer & Smith 1995). The reduction in A was recovered by 95% when water droplets on the leaves were evaporated (Fig. 2). However, when the mist exposure was prolonged to 72 h, a 28% reduction of A was observed after removing water droplets from the leaves for the ‘wettable’ species (Fig. 3). In addition, for the ‘non-wettable’ species, the 72 h mist pretreatment increased A significantly, although no effect of mist was observed for the leaves exposed to 5 min of mist (Fig. 2). These indicate that mist has ‘irreversible’ effects on photosynthesis for both of the ‘wettable’ and ‘non-wettable’ species, if their leaves are exposed to prolonged mist.
In pea, 72 h mist pretreatment resulted in an increase in stomatal conductance (gs; Fig. 3b) and no change in the amount of Rubisco (Fig. 6). This suggests that the increase of A in pea (Fig. 3a) is due to an enhancement of CO2 diffusion through stomata. In a natural environment, the enhancement of CO2 diffusion in mist-treated leaves was attributed to a decrease in the VPD caused by the surface water (Munné-Bosch et al. 1999). However, gs in the mist-treated pea was constantly higher than that in the control leaves when VPD was changed from 500 to 1500 Pa (Fig. 4). This indicates that 72 h mist pretreatment changes the stomatal response to VPD. A previous study showed that gs was significantly higher for the rice plants pretreated to high humidity than those to low humidity when measured below 2.0 kPa of VPD (Kawamitsu, Yoda & Agata 1993). Similarly, the exposure of cuttings of Corylus maxima to fog increased stomatal conductance under similar VPD (Fordham et al. 2001). Our result is consistent with these previous findings.
This increase in gs by the mist pretreatment may be partly induced by an increase in stomatal density in the mist-treated pea (Table 2). On the other hand, water potential of epidermal cells, which was considered to be a cue for stomatal regulation (Monteith 1995; Buckley & Mott 2002), may not cause the increase in gs because of the absence of water absorption by the leaf surface in the mist-treated pea (Fig. 1).
Mist pretreatment of bean leaves resulted in a reduction in stomatal conductance (Fig. 3b) but no change in internal conductance (Table 3). These results suggest that the inhibition of A by the mist pretreatment (Fig. 3a) is at least partly caused by a limitation of diffusion through the stomata. Water absorption from the leaf surface (Fig. 1) might increase water potential more in the epidermal cells than in the guard cells, causing stomatal closure of the mist-treated bean leaves (Fig. 4). Another possibility is that leaching of the ions from the wetted leaves (Tukey 1970) induces a relative decrease in the osmotic potential of the epidermal cells compared with guard cells, resulting in stomatal closure in the wettable species. The fact that mist pretreatment had no effect on internal conductance may be due to the small change in mesophyll anatomy (Table 3).
Perhaps a more important cause of the reduction in A in bean is a decrease in the level of Rubisco, because the amount of both LSU and SSU were reduced remarkably in this species (Fig. 6). The decrease in LSU is consistent with the finding of Ishibashi et al. (1996) that 24 h mist exposure reduced the amount of Rubisco by 45% in bean leaves. The reduction in the amount of Rubisco may be due to an alteration of the balance of Rubisco synthesis and degradation, because Rubisco normally undergoes continuous degradation in mature leaves (Esquivel, Ferreira & Teixeira 1998). The fragments of 42, 31 and 23 kDa in the present study (Fig. 5) are most likely the degradation product of Rubisco LSU. A 36 to 37 kDa fragment was detected as a degradation product of LSU in rice ( Mae et al. 1983), barley (Desimone, Henke & Wagner 1996) and wheat (Ishida et al. 1997), under oxidative stress. Similarly, proteins of 48, 34 and 24 kDa were detected by a Rubisco antibody in leaf extracts from pea plants, which were probably the degradation products of LSU (Romero-Puertas et al. 2002). Therefore, the increased intensities of the 42 and 23 kDa bands (Fig. 7) indicate accumulations of the degradation products. The reduction in LSU and the increase in degradation products of Rubisco in bean indicate that Rubisco turnover is inhibited by mist pretreatment.
How is Rubisco degraded in mist-treated bean leaves? One possibility is that environmental stresses that limit CO2 fixation cause Rubisco fragmentation (Desimone et al. 1996; Ishida et al. 1997), because energy is channelled from CO2 fixation to the production of active oxygen species resulting in photo-inhibition (Moran et al. 1994; Tambussi et al. 2000). In fact, drought stress, which limits CO2 fixation, reduced the amount of Rubisco in the leaves of tobacco (Gunasekera & Berkowitz 1993) and rice (Vu et al. 1998). However, the Rubisco degradation that we observed was probably not caused by photo-inhibition because mist pretreatment did not reduce Fv/Fm (Table 2). A second possibility is that the rapid reduction in CO2 assimilation rate by surface wetness (Figs 2 & 3a) triggers a signal that accelerates senescence. This, in turn, could lead to Rubisco degradation. A decline in photosynthetic rate below a certain threshold triggered the senescence programme (Gan & Amasino 1997). Surface wetness, like other environmental stresses such as drought, extreme temperature, shading and nutrient deficiency, may act as an environmental cue that initiates senescence in ‘wettable’ species.
In conclusion, mist pretreatment had significant ‘irreversible’ effects on photosynthesis for both ‘wettable’ and ‘non-wettable’ species: 72 h mist pretreatment caused a 28% decrease in A in bean, whereas it caused a 22% increase in A in pea. This difference may be due to a difference in the effect of wetness on stomatal conductance (gs) and the amount of Rubisco in these two species. The decrease of A by wetness in bean is due to the decreases in both gs and the amount of Rubisco. On the other hand, the increase in A by wetness in pea is due to the increase in gs only.
This study was supported by a grant from the Ministry of Education, Science, Sports and Culture (no. 12740423) and the Ohara foundation from the Research Institute for Bioresources, Okayama University. The authors thank Dr Y. Yamamoto (Okayama University) for generously donating the antibodies used in this study, and for critical comments on the manuscript. We thank Dr T. Kobayashi (Kagawa University) for constructive comments on the manuscript. The isotope measurements were done at Center for Ecological Research, Kyoto University.