Underwater photosynthesis and respiration in leaves of submerged wetland plants: gas films improve CO2 and O2 exchange


  • Timothy David Colmer,

    1. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia;
    Search for more papers by this author
  • Ole Pedersen

    1. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia;
    2. Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Helsingørsgade 51, DK-3400 Hillerød, Denmark
    Search for more papers by this author

Author for correspondence:
Ole Pedersen
Tel:+45 3532 1900
Fax: +45 3232 1901
Email: opedersen@bi.ku.dk


  • • Many wetland plants have gas films on submerged leaf surfaces. We tested the hypotheses that leaf gas films enhance CO2 uptake for net photosynthesis (PN) during light periods, and enhance O2 uptake for respiration during dark periods.
  • • Leaves of four wetland species that form gas films, and two species that do not, were used. Gas films were also experimentally removed by brushing with 0.05% (v/v) Triton X. Net O2 production in light, or O2 consumption in darkness, was measured at various CO2 and O2 concentrations.
  • • When gas films were removed, O2 uptake in darkness was already diffusion-limited at 20.6 kPa (critical O2 pressure for respiration, COPR≥ 284 mmol O2 m−3), whereas for some leaves with gas films, O2 uptake declined only at approx. 4 kPa (COPR 54 mmol O2 m−3). Gas films also improved CO2 uptake so that, during light periods, underwater PN was enhanced up to sixfold.
  • • Gas films on submerged leaves enable continued gas exchange via stomata and thus bypassing of cuticle resistance, enhancing exchange of O2 and CO2 with the surrounding water, and therefore underwater PN and respiration.


Flooding resulting in complete submergence is a severe stress for terrestrial plants, principally because of impeded underwater gas exchange (Voesenek et al., 2007). Low CO2 availability, and in some cases also low light, restricts photosynthesis in submerged plants (Sand-Jensen, 1989). Leaf adaptations and acclimations in aquatic species enhance access to CO2; examples include morphology and anatomy to reduce boundary-layer and cuticle resistance, and in some species a capacity to utilize inline image to provide more CO2 (Maberly & Spence, 1989; Raven et al., 1985; Madsen & Sand-Jensen, 1991; Frost-Christensen et al., 2003; Frost-Christensen & Floto, 2007). When submerged, heterophyllous species produce aquatic leaf forms with improved rates of underwater net photosynthesis (Sand-Jensen et al., 1992). Leaves of terrestrial species can also acclimate during submergence, so that resistances against CO2 and O2 uptake across the cuticle and epidermis are reduced (Mommer & Visser, 2005; Mommer et al., 2005, 2007).

When leaves of terrestrial plants (or aerial leaves of amphibious species) become submerged, CO2 entry occurs predominantly via diffusion across the cuticle (studied only for leaves without gas films: Frost-Christensen et al., 2003; Mommer & Visser, 2005). However, some emergent wetland species may avoid the problem of high cuticle resistance by the presence of gas films on submerged leaves. Setter et al. (1989) hypothesized that gas films on submerged leaves (termed ‘gas envelopes’ by these authors) provide ‘an interface between the gas and water phases for collection of CO2 and dispersal of O2 during the day or collection of O2 during the night’. The role of leaf gas films has been evaluated in the aeration of partially submerged rice (Raskin & Kende, 1983; Beckett et al., 1988), but little is known about the influence of gas films on CO2 and O2 exchange of completely submerged leaves. For CO2 uptake, data are restricted to one experiment with rice: 14C incorporation was nine- to tenfold higher in leaves with gas films compared with leaves without, but rates of underwater net photosynthesis (PN) were not provided (Raskin & Kende, 1983).

The influence of leaf gas films on underwater PN has not been evaluated for wetland species, except for relative 14C incorporation in rice (described above), nor have the benefits for leaf O2 uptake and respiration (even in rice). Tissues of completely submerged plants rely on O2 uptake from the water column for respiration during the night (Waters et al., 1989; Sand-Jensen et al., 2005; Pedersen et al., 2006), so that enhanced gas exchange with the water column improves internal aeration. Thus leaf gas films might function in a similar way to the well described plastron (gas layer trapped by hairs) found in some aquatic insects (cf. Vogel, 2006). The present study experimentally manipulated gas films on leaves of four emergent wetland species, and also used two species with leaves that do not form gas films, to test the hypothesis that leaf gas films enhance CO2 uptake for PN during light periods, and O2 uptake for respiration during dark periods.

Materials and Methods

Plant materials

We studied six wetland species with leaves of contrasting ability to form gas films when submerged (Table 1). Of these six species, three form prominent gas films on both sides of the leaf; one forms a gas film only on the adaxial leaf surface; and the other two do not form leaf gas films. These gas films can be seen with the naked eye as a silvery covering of the leaf when in direct light. One of the species that does not form leaf gas films (Sparganium emersum L.) is heterophyllous: it develops an aquatic leaf form when submerged. Material used in all experiments was collected during July 2007 from plants growing on the edges of lakes or streams (or completely submerged in the case of the aquatic form of S. emersum) near Hillerød, Denmark. The youngest fully expanded leaf was used, and measurements were taken for the lamina (not including the midrib, if present) midway along the leaf blade. Exceptions were Typha latifolia L., for which tissue was taken three-quarters of the distance towards the leaf tip as the leaves were much thicker at the mid-point; and the aquatic form of S. emersum, for which the midrib area was included as the leaves were narrow and the midrib was not pronounced.

Table 1.  Leaf surface characteristics of the six wetland species studied
SpeciesLeaf sideGas film present (+) or absent (−)Wettability (contact angle, θ°)Stomatal density (number mm−2)
  1. Presence or absence of gas films was evaluated visually; surface wettability was determined using the water droplet contact angle (θ) method (n = 5), surfaces with θ < 110° are considered wettable whereas those with θ > 130° are nonwettable (Brewer & Smith, 1997); stomatal density was assessed on surface impressions of lamina (n = 4) under a microscope. For S. emersum, A, aerial leaf form; S, submerged leaf form. For A. calamus, leaf blades are orientated so that adaxial and abaxial surfaces could not be assigned. Values are means ± SE.

Phalaris arundinacea L.Adaxial+124 ± 6121 ± 9
Abaxial+123 ± 6144 ± 14
Phragmites australis (Cav.) SteudelAdaxial+151 ± 2601 ± 36
Abaxial+154 ± 3937 ± 37
Typha latifolia L.Adaxial+138 ± 8368 ± 34
Abaxial+156 ± 2365 ± 35
Glyceria maxima (Hartman) O. Holmb.Adaxial+153 ± 2109 ± 13
Abaxial108 ± 7 72 ± 12
Acorus calamus L.Side 1 76 ± 3218 ± 6
Side 2 81 ± 5253 ± 20
Sparganium emersum L. (A)Adaxial 48 ± 6133 ± 9
Abaxial 65 ± 4153 ± 7
Sparganium emersum (S)Adaxial 50 ± 5< 1
Abaxial 51 ± 4< 1

Characterization of leaf lamina for selected morphological, chemical and biophysical properties

Leaves were characterized for selected properties (n = 4–5; Tables 1 and 2). Surface wettability was assessed by measuring the contact angle of a 5-mm3 droplet of water on the leaf surface (Adam, 1963; Brewer & Smith, 1997), held flat using double-sided tape. Droplets were applied to the lamina of five replicate leaves of each species, and photographed at ×16 magnification using a horizontally positioned dissecting microscope (Leica MS5, Solms, Germany) and digital camera. Stomatal density was determined using enamel nail polish to make surface impressions of the lamina, and photographing these at ×100 magnification (Olympus SZX12, Tokyo, Japan) and counting stomata on images displayed using computer software (Olympus d-p soft cvi, Tokyo, Japan). Leaf thickness was measured on transverse sections of lamina, using a microscope equipped with a digital camera. Specific leaf area (SLA) of lamina was measured by determining the area (LI-3000, Li-Cor, Lincoln, NE, USA) and dry mass of samples. Total N was determined in oven-dried (65°C), finely ground samples, using a CHN analyser (CHN EA1108, Carlo Erba, Milan, Italy). Chlorophyll was determined in methanol extracts of fresh tissues ground using a pestle and mortar, filtered and analysed with a spectrophotometer (UV-160A, Shimadzu, Japan), using the procedures described by Wellburn (1994).

Table 2.  Selected leaf chemistry and morphological characteristics of the six wetland species studied
SpeciesChlorophyll a (g m−2)Nitrogen (mg m−2)SLA (m2 kg−1)Thickness (µm)Porosity (% volume)
  1. SLA, specific leaf area. Parameters were measured for lamina taken half-way up (except T. latifolia, 2/3-way up) the most recent fully expanded leaf blade. For S. emersum, A, aerial leaf form; S, submerged leaf form. Values are means ± SE (n = 4, with the exception of n = 5 for SLA).

Phalaris arundinacea0.22 ± 0.02 91 ± 619.3 ± 0.8146 ± 26.7 ± 1.2
Phragmites australis0.48 ± 0.01219 ± 1012.5 ± 0.8211 ± 123.6 ± 0.3
Typha latifolia0.63 ± 0.06271 ± 158.6 ± 0.3935 ± 6946.6 ± 2.0
Glyceria maxima0.32 ± 0.03145 ± 1120.0 ± 1.1433 ± 1647.0 ± 1.0
Acorus calamus0.46 ± 0.04202 ± 2518.8 ± 1.6265 ± 2217.0 ± 1.8
Sparganium emersum (A)0.26 ± 0.01121 ± 1319.7 ± 1.6597 ± 9639.7 ± 1.2
Sparganium emersum (S)0.27 ± 0.01 75 ± 142.0 ± 2.4460 ± 3516.9 ± 6.6

Porosity (percentage gas spaces per unit tissue volume) of lamina was measured by determining tissue buoyancy before and after vacuum infiltration of the gas spaces with water (Raskin, 1983), using the equations as modified by Thomson et al. (1990). Triton X (0.05% v/v) was used to remove gas films on the segments of lamina when submerged for the porosity measurements. Care was taken to ensure no external gas was trapped between lamina segments, and to ensure complete infiltration of the tissues during that step in the measurements.

Measurements of underwater net photosynthesis

Lamina segments, from opposite halves of the midrib (if present), or sliced vertically down the middle (if no pronounced midrib), were excised using a razor blade. One half was used as the control (with gas films when present) and the opposite half was used as the treatment in which gas films were removed by pretreatment with Triton X. Segments were 25 mm long. Lamina segments used in the without-gas film treatment were brushed five times, on both sides, with a fine paintbrush soaked in 0.05% (v/v) Triton X in incubation medium (composition given below), then washed for 5 s, three times, in medium without Triton X. This treatment prevented formation of gas films on the lamina surfaces when submerged (for rice leaves cf. Raskin & Kende, 1983), and did not cause any flooding of internal gas spaces (unpublished data). Lamina of two species known not to form gas films were treated in identical fashion, providing a negative control. Experiments commenced within 15 min of excision of lamina segments.

Net photosynthesis underwater, by lamina segments (with or without Triton X pretreatment to remove any gas films), was measured using the method described by Sand-Jensen et al. (1992), with some modifications. The glass bottles used were 25 ml, and a glass bead was added to each bottle to promote mixing as the bottles rotated on the wheel within the illuminated water bath at 20°C. The 25-mm lamina segments tumbled in the rotating flasks (one segment per flask) and addition of one glass bead increased PN, but more beads (up to six) had no further effect. Photosynthetically active radiation (PAR) was 600 µmol m−2 s−1 (measured using a 4π US-SQS/L Walz, Effeltrich, Germany).

The basal incubation medium was based on the general purpose culture medium described by Smart & Barko (1985) and contained (in mol m−3) Ca2+, 0.62; Mg2+, 0.28; Cl, 1.24; inline image, 0.28. The dissolved O2 concentration in the incubation medium was set at 50% of air saturation, by bubbling in 1 : 1 volumes of N2 and air; this procedure was applied to prevent increase in O2 above air-saturated levels during measurements, that might have led to photorespiration and thus decreased PN (for submerged rice leaves cf. Setter et al., 1989). As flasks were incubated in light immediately after adding the lamina segments, and as the segments produce O2 when in light, there was no risk of tissue hypoxia. After the solution had been set with O2 at 50% air saturation, dissolved CO2 treatments were imposed by adding specific concentrations of KHCO3 with the pH always adjusted to pH 7.35 (Stumm & Morgan, 1996) (representing the pH found at the field sites; data not shown), and checked at the end of measurements. K2SO4 was added so that K+ concentration was equal across treatments. Thus, in the two main treatments (50 and 500 mmol m−3 dissolved CO2), KHCO3 was at 0.5 and K2SO4 at 2.6 mol m−3, and KHCO3 was at 5.2 and K2SO4 at 0.25 mol m−3, respectively.

Following incubations of known duration, dissolved O2 concentrations in the solution of the flasks were measured using a Clark-type O2 mini-electrode (OX-500, Unisense A/S, Aarhus, Denmark) connected to a pA meter (PA2000, Unisense). The electrode was calibrated immediately before use in water at atmospheric saturation (20.6 kPa O2) and in anoxic water with dithionite (0 kPa O2). Dissolved O2 concentrations in bottles prepared and incubated in the same way as described above, but without lamina segments, served as blanks. The projected area of each lamina segment was measured using a leaf-area meter (LI-3000), and fresh and dry masses of each segment were also determined. Each species × treatment combination was replicated five times.

Measurements of dark respiration by leaf segments when underwater

Leaf lamina segments were placed in air-saturated incubation medium with the same composition as the 50-mmol m−3 dissolved CO2 treatment used above in PN experiments. In mol m−3: Ca2+, 0.62; Mg2+, 0.28; Cl, 1.24; inline image, 0.28, plus KHCO3 at 0.5 and K2SO4 at 2.6 mol m−3 (pH 7.35 to provide 50 mmol m−3 dissolved CO2; Stumm & Morgan, 1996); this pH was selected as it represents that of the water at the field sites. Measurements were taken using 4-ml glass chambers with a capillary hole in the glass stopper (MicroResp, Unisense) through which an O2 microelectrode (MicroResp EL, Unisense) was inserted. The capillary hole also acted as a vent during closure of the chamber so that pressure within the chamber did not increase. Each chamber contained a small, glass-coated magnetic stir bar. O2 was monitored continuously and the signal from the multichannel pA meter (PA8000, Unisense) was logged every 10 s (ADC16, Pico Technology, St Neots, UK) connected to a laptop computer. The incubation medium was sterile-filtered (0.22 µm) before use. Eight chambers were run simultaneously at 20°C in darkness. After each experiment, the volume (±0.001 ml) in each chamber was determined to account for the slightly different tissue volume in each replicate. The experiments were run for 3–5 h, depending on how quickly O2 was depleted from the chambers. The projected area of each lamina segment was measured using a leaf-area meter (LI-3000), and fresh and dry masses of each segment were also determined. Each species × treatment combination was replicated four times.

Data analyses

graphpad prism 4.0 (GraphPad Software Inc., San Diego, CA, USA) was used to fit the Jassby & Platt (1976) model to the CO2-response curve for Phragmites australis and to describe the relationship between respiration rate and external O2 concentration. The critical O2 pressure for respiration (COPR) is the O2 concentration below which O2 uptake by a tissue first declines (Berry & Norris, 1949). In the present study we chose a 10% reduction as being a significant decline in O2 uptake rate. graphpad prism 4.0 was also used for anova and t-tests to compare means.


Gas films enhance underwater net photosynthesis in leaves of several wetland species at low or high external CO2

Leaves of Phalaris arundinacea, P. australis and T. latifolia had gas films on both sides when submerged; Glyceria maxima possessed a gas film on the adaxial, but not the abaxial side; and Acorus calamus and S. emersum (both aerial and submerged leaf forms) did not form gas films (Table 1). When excised shoots were submerged in an aquarium containing water with 275 mmol CO2 m−3 (20°C; PAR 200 µmol m−2 s−1, 16 h light : 8 h dark), gas films persisted for at least 2 wk on leaves of the four species tested (experiment terminated at 2 wk).

Underwater PN was higher, by up to six times at both low and high external CO2 concentrations, in leaves with gas films compared with the same species after gas films were removed (Fig. 1a,b). At 50 mmol m−3 CO2, presence of a gas film increased PN only 1.5 times in G. maxima, whereas PN increased six times in P. arundinacea. At 500 mmol m−3 CO2, gas films increased PN 3.1 times in P. arundinacea, but 6.8 times in P. australis. Thus gas films had a large stimulatory effect on underwater PN by leaves of four wetland species, but the magnitude was variable for the different species. Measurements were conducted at 50 and 500 mmol m−3 external CO2, so that the influence of external CO2 could be assessed, and these concentrations also represent ranges likely to occur in the natural local habitats of the species used (Sand-Jensen et al., 1992; unpublished data from collection sites of the species used in the present experiments). For all species and Triton X treatment combinations, the tenfold increase in external dissolved CO2 resulted in increased PN; increases ranging from 2.1- to 13.1-fold (Fig. 1a,b).

Figure 1.

Rates of underwater net photosynthesis (PN) in leaf lamina segments of six wetland plant species, with gas films (when present for particular species) (control, open bars), or after pretreatment with 0.05% (v/v) Triton X to remove any gas films (closed bars) at (a) 50, (b) 500 mmol m−3 free CO2. PAR was 600 µmol m−2 s−1; temperature was 20°C. Values are means ± SE (n = 5).

Species that do not form leaf gas films were included in the experiments as negative controls. The pretreatment of brushing the leaf surface with 0.05% (v/v) Triton X had no effect in the case of A. calamus, whereas for both leaf forms of S. emersum, control leaves had PN rates 10–30% higher than for leaves after treatment with Triton X (Fig. 1a,b). These responses to Triton X treatment in S. emersum were small, in comparison with the much larger decreases in PN when gas films were removed from leaves of the four species with these films.

A feature of the leaf surfaces with gas films when submerged was that these were all classified as ‘nonwettable’ (Table 1), based on the contact angles of water droplets applied to lamina being ≥ 130° (cf. Brewer & Smith, 1997). By contrast, contact angles for leaf surfaces that do not form gas films were either equal to the 110° considered to be the upper limit for wettable surfaces (e.g. abaxial surface of G. maxima), or considerably less than this value (A. calamus and S. emersum). Other features of the leaves of species that formed gas films were diverse (Table 2); as examples: tissue porosity ranged from as low as 3.6% in P. australis to approx. 47% in T. latifolia and G. maxima; leaf thickness varied 6.4-fold; SLA varied 2.3-fold; leaf N and chlorophyll a varied approx. 2.9-fold. So, although the photosynthetic capacities of these diverse materials varied (Fig. 1), and rates of PN were correlated with some parameters (e.g. positively with leaf N when external CO2 was 500 mmol m−3; Fig. 1; Table 2, r2 = 0.33–0.40), gas films had significant beneficial effects on PN in these diverse leaf materials.

CO2-response curves for underwater net photosynthesis in leaves of P. australis with or without gas films

CO2-response curves were established for leaves of P. australis with or without gas films (Fig. 2). PN was enhanced 5.6–7.9-fold by the presence of gas films, across a 100-fold range of external CO2 concentrations from 20 to 2000 mmol CO2 m−3. The model developed by Jassby & Platt (1976) was fitted to the data (r2 = 0.955–0.964) and the apparent Pmax for leaves with gas films was seven times higher than that of leaves without gas films. For this experiment, however, we had hypothesized that, with high external CO2, PN of the leaf segments without gas films would approach the rates of those with gas films; high external CO2 might be expected to overcome diffusion limitations. In the case of O2 uptake during darkness, the rates of control and Triton X-treated leaf segments did converge (Fig. 3); these O2-uptake data are discussed below. We tested the possibility that high inline image (40 mol m−3 in the solution with 2000 mmol CO2 m−3) might have adverse effects when in contact with the leaf segments (after gas films were removed). PN in the bicarbonate buffer was compared with that in a 2 mol m−3 MES buffer at pH 6.0, so that at the same free CO2, inline image was 1 mol m−3. There was no difference in PN between leaves in the two buffers (data not shown). In addition, we also tested whether use of a different detergent (Tween 20) to remove gas films produced the same effect as Triton X, and found no difference (tested at 2000 mmol CO2 m−3; data not shown). Further interpretation of the differences in Pmax between leaves with and without gas films (Fig. 2) is limited by a lack of data on CO2 permeability of the cuticle, whether stomata close or remain open when the gas films are removed, and by possible changes in pH and therefore free CO2 in the unstirred boundary layers. Notwithstanding this enigmatic result, the data in Fig. 2 clearly demonstrate the beneficial influence on PN of gas films for submerged leaves over a wide range of external CO2 concentrations.

Figure 2.

Responses to free CO2 concentration of underwater net photosynthesis (PN) in leaf lamina segments of Phragmites australis with (□) or without (inline image) gas films. Gas films were removed by pretreatment with 0.05% (v/v) Triton X. PAR was 600 µmol m−2 s−1; temperature was 20°C. Apparent Pmax for leaf segments with gas films was 6199 nmol m−2 s−1; for those without gas films, 891 nmol m−2 s−1. Values are means ± SE (n = 5).

Figure 3.

O2-uptake rates at decreasing external O2 concentrations in leaf lamina segments of Phragmites australis (a) with (control); (b) without gas films. Gas films were removed by pretreatment with 0.05% (v/v) Triton X. Measurements were taken in darkness at 20°C. Air-saturated water at 20°C = 284 mmol m−3. Inserts show an example of recordings for individual replicates over the full concentration range: x- and y-axes of inserts are identical to main axes. Values are means ± SE (n = 4).

Gas films enhance dark respiration by leaves when underwater in darkness

The presence of gas films enhanced leaf O2 uptake when submerged in darkness. With the gas films present in P. australis, the rate of O2 uptake was virtually independent of the surrounding O2 concentration down to 60 mmol O2 m−3, after which diffusion limitation set in resulting in a steep decline in O2 uptake (Fig. 3). By contrast, when the gas films were removed, the rate of O2 uptake was diffusion-limited even at air saturation (284 mmol O2 m−3; Fig. 3). The maximum (nondiffusion-limited) rate of respiration (Rmax) was apparent for leaf segments of the four emergent species with gas films, and for the submerged leaf type of S. emersum, whereas in the two emergent leaf types without gas films (A. calamus and the aerial form of S. emersum) Rmax might not have been reached even at air saturation (Fig. 4a). Rmax was approx. 400 nmol O2 m−2 s−1 in P. arundinacea, P. australis and T. latifolia, and therefore was not closely related to leaf N (Table 2). Rmax in G. maxima, and in the aquatic leaf form of S. emersum, was approx. 20% lower than the rates in P. arundinacea, P. australis and T. latifolia (Fig. 4a).

Figure 4.

(a) Maximum rates of O2 uptake (Rmax); (b) critical O2 pressure (COPR) for O2 consumption by leaf lamina segments of six wetland plant species. Measurements were taken in darkness at 20˚C. Rmax is the nondiffusion-limited rate of O2 uptake. COPR, external O2 concentration at which O2 uptake first declines by 10% from Rmax. Rmax and COPR were calculated from the Jassby & Platt (1976) function fitted to each replicate data set. Data shown only for control leaves (no Triton X pretreatment, gas film present; Table 1) as Rmax was not reached in air-saturated solution for several species when gas films were absent, so Rmax and thus COPR for Triton X-treated leaves is not available (COPR ≥ 284 mmol O2 m−3). For Sparganium emersum: A, aerial leaf form; S, submerged leaf form. Values are means ± SE (n = 4; except S. emersum (A) Triton X-treated, n = 2; S. emersum (S), n = 3); NA, not available.

Critical oxygen pressure for O2 uptake (COPR) in leaves of two of the four terrestrial wetland plants with gas films was comparable with that (approx. 65 mmol O2 m−3) of the aquatic leaves of S. emersum, and although COPR was not as low as this in the other two species with gas films, COPR was still well below air saturation (Fig. 4b). By contrast, for the two emergent leaf types without gas films when submerged, COPR ≥ 284 mmol O2 m−3 (Fig. 4b). Thus O2 uptake by these leaves naturally without gas films, and for all leaves with gas films removed experimentally (data for P. australis, Fig. 3), was diffusion-limited even near air saturation.

The beneficial effect of leaf gas films on O2 uptake is also demonstrated by calculation of the apparent resistance to O2 uptake for the control and Triton X-treated leaf segments (Fig. 5). Removal of gas films from leaves of P. arundinacea and P. australis increased resistance to O2 uptake by approx. three- to fourfold. The adverse effect of gas film removal was less pronounced in species with leaves of high porosity, such as G. maxima and T. latifolia (Fig. 5). When gas films were removed, leaves with high tissue porosity showed a lower resistance to O2 uptake than the species with low porosity (Fig. 5), possibly because of a greater external liquid–air interface area at the cut surfaces of high-porosity leaf tissues. For species that had no leaf gas film when submerged (A. calamus and S. emersum), Triton X treatment had no, or little, effect on O2 uptake compared with controls, as reflected in the comparable values of resistance to O2 uptake (Fig. 5).

Figure 5.

Apparent resistance (r) to O2 uptake by leaf lamina segments of six wetland plant species. Measurements were taken in darkness at 20°C. Apparent resistance to O2 diffusion during uptake (r) = 1/conductance = 1/slope of a linear regression fitted to O2 uptake data in the 2–30-mmol O2 m−3 range for each replicate data set. Control, open bars; Triton X-pretreated, closed bars. For Sparganium emersum: A, aerial leaf form; S, submerged leaf form. Values are means ± SE (n = 4; except S. emersum (A) Triton X-treated, n = 2; S. emersum (S), n = 3); *, Significant difference (P < 0.05) between the treatments for a pairwise comparison within each species.


The present data support the hypothesis that gas films on the surface of submerged leaves enhance gas exchange with the surrounding water, and thus underwater PN (CO2 uptake during the day) and respiration (O2 uptake during the night). For submerged leaves of P. australis, gas films enhanced PN 5.6- to 7.9-fold across a 100-fold range of external CO2 concentrations (Fig. 2), and decreased the apparent resistance for O2 uptake to 21% of that when the gas film was removed (Fig. 5), so that tissue COPR was significantly lowered. The likely mechanisms by which gas films enhance leaf gas exchange with the water column are: an increased gas–liquid interface between leaves and floodwater; and continued gas-phase exchange of CO2 and O2 via stomata, thus by-passing the cuticle resistance. This situation of gas exchange continuing via the stomata of submerged leaves with gas films contrasts with the situation proposed for leaves of terrestrial species lacking gas films when submerged, for which it has been suggested that stomata close and exchange with the water column occurs primarily across the cuticle (Mommer & Visser, 2005; Mommer et al., 2005, 2007).

The presence of gas films on leaves of certain terrestrial wetland plants appears to have been overlooked in studies of underwater PN. We observed that touching the leaf surface reduced the gas films, so that these might have been less obvious, or even absent, in earlier studies. This sensitivity of gas films to leaf handling might explain why earlier recorded rates of underwater PN in G. maxima and P. australis, at similar CO2 concentrations, were two- and fivefold lower (Sand-Jensen et al., 1992), respectively, compared with the present study. In contrast to the situation for ‘wild’ wetland species, gas films have been noted on submerged leaves of rice (Raskin & Kende, 1983; Setter et al., 1989), and the significance for aeration of partially submerged rice has been evaluated (Raskin & Kende, 1983; Beckett et al., 1988). However, for completely submerged leaves of rice, although the relative differences in 14CO2 incorporation between rice leaf segments with or without gas films when underwater were documented as a nine- to tenfold difference (Raskin & Kende, 1983), the actual photosynthetic rates were not quantified and the influence on O2 uptake for respiration was not studied.

Leaf gas exchange is greatly impeded by submergence because of the 10 000-fold lower diffusion coefficients of O2 and CO2 in water compared with air (Voesenek et al., 2007). Even thin films of water on leaves of terrestrial species, resulting from dewfall or rainfall, can greatly impede leaf gas exchange (Smith & McClean, 1989). Avoidance of thin water films on leaves of terrestrial plants has been related to the water repellency of the leaf surfaces (Smith & McClean, 1989; Brewer & Smith, 1997). In a survey of leaf water repellency of 200 diverse plant species, Neinhuis & Barthlott (1997) noted that ‘most marsh and water plants were found to be water-repellent’. The present study indicates that leaf water repellency also appears to be associated with the development of surface gas films when submerged (Table 1), adding to the previously recognized functional roles of differences in wettability of leaves (other functions are discussed by Neinhuis & Barthlott, 1997; Brewer & Smith, 1997). Leaf wettability is determined by the micro- and nanostructures of the surface, as well as wax crystals (Wagner et al., 2003; Bhushan & Jung, 2006), so these structural features presumably determine whether a gas film is present when leaves are submerged.

The rates of underwater PN during light periods (Fig. 1), and for dark respiration (Fig. 4a), allow for an estimate of carbon balance of leaf tissues. With the gas films present, leaves of P. australis would have a positive C balance of 15 and 137 mmol m−2 d−1 at 50 and 500 mmol m−3 CO2, respectively, assuming 12 h light and 12 h dark. By contrast, without the gas films the C balance becomes negative (−4 mmol C m−2 d−1) at low CO2, but was positive (9 mmol C m−2 d−1) at high CO2. Data for whole-plant C balances when fully submerged are not available for the species studied; however, the leaf-based estimates above indicate the potential importance of gas films for the C balance of submerged terrestrial plants.

In conclusion, gas films on leaves of submerged wetland plants enable continued gas exchange via stomata and thus bypassing of cuticle resistance, enhancing exchange of O2 and CO2 with the surrounding water. Resistance (= 1/conductance) of the cuticle to CO2 movement can be extremely high (e.g. grape leaves; Boyer et al., 1997), yet acclimation by aquatic leaves can reduce cuticle resistance several-fold and thus enhance underwater PN (e.g. in Rumex palustris, with no gas films, by 69-fold at external CO2 of 250 mmol m−3; Mommer et al., 2006). It is of significance to the present study that rates of underwater PN in terrestrial leaves with gas films were similar to those of the submergence-acclimated leaves of R. palustris (Mommer et al., 2006) and the aquatic leaf form of S. emersum (Fig. 1). Thus, during submergence, gas films on leaves enhance the photosynthesis and respiration of some wetland plants.


We thank B. Kjøller for assistance with tissue N analyses, and Professors K. Sand-Jensen and H. Greenway for useful critical comments on a draft of the manuscript. The research was funded by CLEAR Lake Restoration, a Villum Kann Rasmussen centre of excellence.