A study of stomatal mechanics using the cell pressure probe


  • P. J. Franks,

    1. Environmental Biology Group, Research School of Biological Sciences, Australian National University, and ,
    Search for more papers by this author
    • Present address: Department of Tropical Plant Sciences, James Cook University, PO Box 6811, Cairns QLD 4870, Australia.

  • 1,2 I. R. Cowan,

    1. Environmental Biology Group, Research School of Biological Sciences, Australian National University, and ,
    Search for more papers by this author
  • G. D. Farquhar

    1. Environmental Biology Group, Research School of Biological Sciences, Australian National University, and ,
    2. Cooperative Research Centre for Plant Science, GPO Box 475, Canberra, ACT 2601, Australia
    Search for more papers by this author

P. J. Franks
Fax: 61 70 42 1284; e-mail: Peter.Franks@jcu.edu.au


The relationship between stomatal aperture (a) and guard cell pressure (Pg) was measured directly in four different species (Vicia faba, Tradescantia virginiana, Ginkgo biloba and Nephrolepis exaltata) using a special cell pressure probe technique. The effect of epidermal turgor (Pep) on this relationship was also measured in T. virginiana. The relationship was sigmoidal for V. faba and T. virginiana, but entirely convex for G. biloba and N. exaltata. Epidermal turgor was found to have a pronounced closing effect on stomata of T. virginiana. Maximum aperture with full epidermal turgor (0·92 MPa) was about half that with zero epidermal turgor. Also, with full epidermal turgor stomata of T. virginiana did not begin to open until Pg was more than 1·25 MPa. These characteristics were used to develop an expression for a as a function of Pg and Pep. Results for the different species are compared and discussed in terms of possible advantages and limitations of water economy.


Stomatal movements are largely determined by the mechanical properties of the guard cells and the epidermal cells with which they interact. It has been known for some time (von Möhl, 1856) that stomata open and close due to changes in guard cell pressure, and that the pressure in epidermal cells limits the extent to which stomata can open. These initial observations form the basis of our current understanding of stomatal movements, and relatively little information has been added to this particular aspect of stomatal physiology in recent times. This is despite the fact that over the last few decades the wider study of stomatal movements has developed into a very large field within botanical science.

The first interpretation of these pressure-driven stomatal movements arose from the study of guard cell anatomical features. It became obvious that, for guard cells to bend away from each other with increasing turgor, rather than press more firmly together like normal cells, their structural characteristics had to be very different from those of normal cells. Early anatomical investigations (Schwendener, 1881; Haberlandt, 1884) showed that guard cells had special patterns of wall thickening that, in theory, resulted in differential expansion as guard cells inflated, and therefore bending in the characteristic manner. These studies also revealed that characteristic patterns of thickening were associated with particular modes of guard cell movement in particular groups of plants.

Until recently, little was understood about the actual forces involved in guard cell movement. Using principles of mechanical theory, DeMichele & Sharpe (1973) hypothesized that epidermal cells would have considerable mechanical advantage over guard cells, in other words considerably less turgor would be required in epidermal cells to close stomata than in guard cells to open them. Meidner & Edwards (1975) and Edwards et al. (1976) attempted to quantify this by directly measuring pressures in guard cells of Tradescantia virginiana with a cell pressure probe. Their results contributed substantially to an understanding of stomatal mechanics and demonstrated quantitatively the effect of subsidiary cell turgor on stomatal aperture in T. virginiana. However, some discrepancies remain unresolved. In particular, the pressure of 0·7 MPa reported by Meidner & Edwards (1975) as being sufficient to fully open T. virginiana guard cells was much less than what might be estimated from measurements of guard cell osmotic pressures (Bearce & Kohl 1970; MacRobbie 1980). As has been noted, however, low epidermal turgor promotes wider stomatal opening at any given guard cell pressure (Meidner & Edwards 1996). There have been no further studies involving the direct measurement of guard cell turgor and stomatal aperture at different epidermal cell turgors. Furthermore, there are no data of this nature for species other than T. virginiana.

In this paper we use a previously described cell pressure probe technique (Franks et al. 1995) to explore two important aspect of the mechanics of stomatal movement. We first obtain and compare the relationships between stomatal aperture and guard cell pressure in three different species (Vicia faba, Nephrolepis exaltata and Ginkgo biloba). We then examine the mechanical advantage in Tradescantia virginiana by measuring guard cell pressure and aperture at different epidermal cell turgors. This information is used to develop a simple mathematical description of stomatal aperture as a function of pressure in guard cells and epidermal cells.


Plant material

Tradescantia virginiana L. is a perennial, mesophytic, terrestrial herb and Vicia faba L. is a climbing, mesophytic annual herb. Nephrolepis exaltata (L.) Schott is a terrestrial or epiphytic fern usually occurring in mesic, shady environments but sometimes in drier habitats. Ginkgo biloba L. is a deciduous tree from temperate regions. T. virginiana and N. exaltata plants were cloned by dividing parent plants. V. faba and G. biloba were grown from seed. G. biloba was grown in 10 dm3 pots under 50% shade cloth in the open, and plants were 4 years old at the time of these experiments. The other species were grown in 3 dm3 pots under 50% shadecloth in the glasshouse (day/night air temperature 30/25 °C, relative humidity 80%). Soil used was 5:2:2:1 compost:sand:peat:perlite incorporating a controlled release fertilizer (Osmocote Plus, Grace-Sierra Pty Ltd, Castle Hill, Australia). Plants were well watered at all times.

Pressure probe measurements

The pressure probe apparatus and technique for obtaining guard cell aperture/pressure relationships have been described in detail elsewhere (Franks et al. 1995). The pressure probe used here was similar in principle to the conventional type (Husken et al. 1978; Steudle & Tyerman 1983; Murphy & Smith 1994), but with modifications enabling it to operate at very high pressures (up to about 6·0 MPa). In particular the probe incorporated high-pressure o-ring seals around the plunger shaft and the glass capillary. The relationship between guard cell hydrostatic pressure and stomatal aperture was obtained by simultaneously pressurizing both guard cells by injecting silicone oil via the pressure probe.

Species comparisons

Epidermes were peeled from the abaxial surface of newly mature leaves of each species using a dissecting procedure that maintained epidermal cells intact. This procedure was carried out under a dissecting microscope. A piece of leaf (5 × 10−3 m by 15 × 10−3 m) was cut and secured to the microscope stage with the abaxial surface facing upwards. Using two fine-pointed forceps, the abaxial epidermis was lifted at one end and then peeled off while maintaining an angle of less than 20° with the mesophyll. The peels were incubated for 1 h in the dark in distilled water buffered at pH 6·5 with 25 mol m–3 MES [2-(N-morpholino) ethanesulphonic acid]. NaOH was used to adjust the buffer pH. The peels were then mounted cuticle-up in a well slide and fixed in place with two small drops of ‘valap’ (1:1:1 vaseline: lanolin: paraffin). The well was then filled with incubation medium and guard cell aperture versus pressure characteristics measured using the technique described in Franks et al. (1995). Epidermal turgor was subsequently measured with the same probe using the standard technique (described in the review by Steudle 1993).

Effect of epidermal turgor

In this experiment, guard cell aperture versus pressure characteristics were obtained at two different epidermal turgors: maximum and zero. Epidermes were peeled from the abaxial surface of newly mature T. virginiana leaves. Half of these were incubated for 1 h in the dark in distilled water buffered at pH 6·5 with 25 mol m–3 MES. These conditions are assumed to have promoted maximum turgor in epidermal cells. The other half were also incubated for 1 h in the dark in distilled water buffered at pH 6·5 with 25 mol m–3 MES, but containing 400 mol m–3 mannitol. This concentration of mannitol, giving the bathing solution an osmotic pressure of about 1·0 MPa, was sufficient to completely remove turgor and bring epidermal cells to a state of incipient plasmolysis. Guard cell aperture versus pressure characteristics and, subsequently, epidermal pressures were obtained for each separate treatment by the methods used for species comparisons.


Plots of stomatal aperture against guard cell pressure (Pg) for given epidermal turgor (Pep) are shown in Figs 1–3 for V. faba, N. exaltata and G. biloba, respectively. The response in V. faba is sigmoidal; in other words, the rate of change of aperture with guard cell pressure first increases then decreases. This sigmoidal characteristic is very similar to that of T. virginiana (Franks et al. 1995; see also Figs 4 & 5). Under these conditions, near-maximum aperture for stomata of V. faba occurred with guard cell turgors of between 4·0 and 5·0 MPa.

Figure 1.

. Plot of stomatal aperture against guard cell hydrostatic pressure (Pg) for Vicia faba with an epidermal turgor (Pep) of 0·62 MPa. The error bar shows the standard error for three replicate stomata of similar dimensions on the same epidermal peel at that pressure. The epidermal peel was prepared as in Franks et al. (1995), i.e. bathed in distilled water that was buffered at pH 6·5 with 25 mol m–3 MES. l, Pg increasing; n, Pg decreasing.

Figure 4.

. Plot of stomatal aperture against guard cell hydrostatic pressure, Pg, for two different T. virginiana guard cell pairs, one experiencing full epidermal turgor (Pep = 0·92 MPa), and the other experiencing zero epidermal turgor (Pep = 0·00 MPa). The stomata were on different epidermal peels from different leaves of the same plant. The control (0·92 MPa epidermal turgor) was incubated for 1 h in the dark in distilled water buffered at pH 6·5 with 25 mol m–3 MES. The other peel (zero epidermal turgor) was incubated for 1 h in the dark in distilled water buffered at pH 6·5 with 25 mol m–3 MES, but also containing 400 mol m–3 mannitol to remove epidermal turgor. All procedures were at 21 °C. The two guard cell pairs had the same length at zero aperture (70 μm). The closing force applied by subsidiary + epidermal cells is very pronounced. ▴,l, Pg increasing; n,▾, Pg decreasing.

Figure 5.

. A repeat of the experiment shown in Fig. 4, but using stomata on epidermal peels taken from the same leaf. The results are consistent with those shown in Fig. 4 for stomata from different leaves. Guard cells in each plot had the same length at zero aperture (70 μm). Error bars for three replicate stomata in each treatment are shown for the guard cell pressure of 3·6 MPa. Note that for zero epidermal turgor and zero guard cell pressure stomata were already slightly open. ▴,l, Pg increasing; n,▾, Pg decreasing.

The aperture versus pressure characteristics for N. exaltata and G. biloba appear similar to each other but they are not of the sigmoidal nature typical of V. faba. Instead, they are characterized by only a decreasing rate of change of aperture with guard cell pressure. Curiously, the epidermal turgor in N. exaltata was quite low (0·08 MPa), and this may partly explain why at guard cell pressures close to zero the stomata of N. exaltata remain partly open (see below). Epidermal turgor could not be measured in G. biloba as the epidermal cells were too tough to allow penetration by the glass probe. However, we believe that epidermal turgor was quite high in G. biloba when these data were collected as there was little deflection of the epidermal cell walls when the probe tip was pushed hard against them. Alternatively, cell walls may simply have been very rigid. Only one complete set of data was obtained for G. biloba due to difficulty in preparing suitable epidermal peels and inserting the probe into guard cells that were slightly obscured by subsidiary cells.

The results of experiments with T. virginiana, involving the manipulation of epidermal cell turgor, are shown in Figs 4 and 5. The data for different epidermal cell turgors in Fig. 4 were collected from stomata on different leaves. These results are very similar to those in Fig. 5 showing data collected for guard cells on the same leaf. Evidently the influence of epidermal cell turgor on stomatal aperture is consistent with previous observations (von Möhl 1856; Heath 1938; Spence et al. 1983; Meidner & Edwards 1975; Edwards et al. 1976). At full epidermal turgor (0·9 MPa), maximum stomatal aperture in T. virginiana was half that at zero epidermal turgor. Also, at full epidermal turgor the stomatal pores in T. virginiana did not begin to open until guard cell pressure was greater than 1·25 MPa. At zero epidermal turgor there was no guard cell threshold pressure for stomatal opening; in fact stomata remained partially open (1–2 μm) when epidermal and guard cell pressures were zero.

To illustrate more clearly the effect of epidermal cell turgor on stomatal aperture in T. virginiana, the reduction in aperture at full epidermal turgor, expressed as a percentage, is shown in Fig. 6. It can be seen that total closure is maintained if guard cell pressure is less than about 1·25 MPa. At higher pressure, the reduction in aperture is diminished, becoming constant at about 50% with pressures greater than 4 MPa. It is improbable that the force exerted by the epidermal cell on the dorsal wall of the guard cell diminishes as guard cell pressure is increased. Therefore its decreasing effect on aperture can be ascribed to the increasing rigidity of the distended guard cell.

Figure 6.

. The effect of full epidermal turgor (0·92 MPa) on stomatal aperture in T. virginiana, using data from Fig. 4. Maximum aperture is reduced by about 50%. At lower guard cell pressure, aperture is reduced even more. When guard cell pressure is below 1·25 MPa, epidermal turgor causes 100% stomatal closure.


The similarities between T. virginiana and V. faba are to be expected. Both are herbaceous plants with guard cells of similar shape and dimensions (despite the fact that T. virginiana is a monocotyledon and V. faba is a dicotyledon). Both being angiosperms, their guard cells are composed of the same materials, i.e. mainly cellulose and pectins. Also, a comparison of the characteristic guard cell wall conformations during stomatal opening reveals close similarities. T. virginiana, being a member of the Commelinaceae, has ‘Commelina type’ guard cell inflation characteristics (Allaway & Milthorpe 1976). The drawings by Raschke & Dickerson (1972) for V. faba show similar characteristics. Furthermore, similarities in the stomatal mechanics of these two species is suggested by their similar stomatal geometry. A biophysical study by Wu et al. (1985) concluded that the mechanical interaction between guard cells and adjacent cells during stomatal movements is determined by their geometric relationship. Transverse sections of stomata in T. virginiana and V. faba show that their guard cell/subsidiary cell or guard cell/epidermal cell complexes are geometrically similar.

The two phases of stomatal opening in V. faba and T virginiana may be due to changing elastic properties of the guard cell walls during inflation (Wu & Sharpe 1979). This is certainly plausible, but requires more supportive experimental evidence. The most detailed studies on the arrangement and properties of guard cell micellae (the cellulose chains which make up the microfibrils) were those of Ziegenspeck (1938, 1955a,b). However, Wu & Sharpe (1979) noted that there was no theoretical understanding of the behaviour of microfibrils in the walls of guard cells during stretching, and at present this still seems to be the case. Furthermore, it is not known whether the arrangement of microfibrils in stomata is consistent, and the precise nature of their role in stomatal mechanics is not firmly established. For example, Mishkind et al. (1981) reported that the arrangement of microfibrils in the walls of Cyperus esculentus was subject to environmental modification, being axially arranged in glasshouse plants, and radially arranged in field-grown plants. This would be an interesting system with which to explore the function of microfibrils in stomatal mechanics, but unfortunately, the guard cells in C. esculentus (≈20 μm in length) are too small for pressure probe work.

In the past there has been some disagreement about the relative importance of wall microfibril arrangement and wall thickenings in causing guard cells to bend and create the stomatal pore during inflation. Differential guard cell wall thickenings, seen in some of the earliest drawings of guard cells (Haberlandt 1884), have long been proposed as a means by which the differential expansion of guard cell walls takes place. However, Aylor et al. (1973) insisted that it was the radial arrangement of microfibrils in the walls that provided guard cells with all the mechanical properties necessary for guard cell function. It is likely that wall thickenings and microfibril arrangement both contribute to the mode of guard cell deformation during inflation. It may be that the chemical composition of guard cell walls is another important factor, and that characteristics of inflation are related to the relative abundance, for example, of cellulose, pectin and lignin.

The results for Ginkgo biloba (Fig. 3) were distinctly different from those for V. faba or T. virginiana at high epidermal turgor, showing only what could be described as a single-phased relationship between guard cell aperture and pressure. This characteristic results in a comparatively much sharper increase in aperture (as a fraction of total aperture) with pressure at the point of stomatal opening. The implications of this, in terms of the gas exchange characteristics of G. biloba, are that for any condition leading to a decrease in guard cell pressure, especially at lower pressures, stomata of G. biloba will show a greater proportional decrease in stomatal aperture (and, presumably, average stomatal conductance) than V. faba or T. virginiana. Under certain conditions, this could be interpreted as being an advantage, in that stomata of G. biloba could be more sensitive to fluctuations in leaf water potential, and therefore better regulators of plant water status. Little can be inferred from the observations on N. exaltata (Fig. 2). It seems likely that the failure of the stomata to close at zero guard cell pressure and perhaps, also, the lack of any tendency of the characteristic towards a sigmoidal form are related to the surprisingly small epidermal cell pressure. Whether the small pressure is an experimental artifact is an open question.

Figure 3.

. Plot of stomatal aperture against guard cell hydrostatic pressure (Pg) for the gymnosperm Ginkgo biloba. Epidermal turgor could not be measured as the probe could not be made to penetrate the seemingly tough cell walls.

Figure 2.

. Plot of stomatal aperture against guard cell hydrostatic pressure (Pg) for the fern Nephrolepis exaltata, with an epidermal turgor of 0·08 MPa. The standard error for three replicate stomata of similar dimensions on the same epidermal peel at Pg = 3·6 MPa is smaller than the symbol. l, Pg increasing; n, Pg decreasing.

The reason for the differences in guard cell inflation characteristics described above may have something to do with a fundamental difference in the guard cell wall materials of these species. Lignin has been reported as occurring abundantly in the guard cell walls of gymnosperms and pteridophytes, but, surprisingly, it is rare in the guard cells of angiosperms (Ziegler 1987). Lignin is present most abundantly in wood, where it adds strength to vessels, fibres and tracheids by making them more resistant to tension and compression. It could be assumed, then, that heavily lignified guard cell walls would be much more resistant to deformation than those composed mainly of cellulose. This may be why guard cells of gymnosperms and pteridophytes show a much smaller range of movement than those found in angiosperms. It is possible that the ability to operate over a much wider range of stomatal apertures contributes to the great adaptability of angiosperms. Clearly, there is the potential for a much better understanding of these groups of plants simply by looking more closely at the physical and chemical properties of their guard cell walls.

The results in Figs 4 and 5 confirm that epidermal turgor in T. virginiana exerts a considerable force on the guard cells, with the potential to cause complete stomatal closure at low guard cell pressures and up to 50% reduction in aperture at guard cell pressures above 4·0 MPa. In a recent communication to Plant, Cell and Environment, Meidner & Edwards (1996) discuss the discrepancy between their results with T. virginiana (Meidner & Edwards 1975) and the results of Franks et al. 1995. They required only 0·7 MPa to inflate guard cells to near maximum aperture, but Franks et al. (1995) showed that, with an epidermal turgor of 0·41 MPa, a guard cell pressure of 4·0 MPa was required to open T. virginiana stomata to near maximum aperture. Meidner & Edwards (1996) suggested that this is because in their experiments the removal of epidermes from transpiring leaves, followed by immersion in liquid paraffin, resulted in there being zero epidermal turgor and no epidermal mechanical advantage during the course of their measurements. However, the results in Figs 4 and 5 show that, when epidermal turgor is zero, a guard cell pressure of 0·7 MPa only opens stomata to an aperture of about 8 μm, less than one-third of that reported by Meidner & Edwards (1975). This discrepancy seems surprisingly large if it were due simply to differences in the biological material, and requires further investigation.

A further discrepancy not discussed by Meidner & Edwards (1996) is that of the pressure required to initiate stomatal opening. Meidner & Edwards (1975) were unable to initiate opening in closed stomata with a pressure of 1·0 MPa (the maximum for their system). In our experiments, full epidermal turgor (0·9 MPa) demanded 1·25 MPa to initiate stomatal opening in T. virginiana, but at zero epidermal turgor no pressure was required to open stomata as they were in fact already slightly open (apertures 1–2 μm). One possible explanation for this is that the probe used by Meidner & Edwards (1975) was particularly susceptible to blockage by cytoplasmic contents when inserted into the guard cells of closed stomata. We found no evidence of a pressure ‘hurdle’ between closed and slightly open stomata. Aperture increased with pressure in a smooth manner after sufficient pressure had been generated in guard cells to push back the subsidiary cells.

Sigmoidal functions were fitted to the data of Fig. 4 and are plotted in Fig. 7. The arrows represent the extent to which apertures are reduced by epidermal turgor. In theory, all the possible combinations of guard cell pressure and aperture lie within the region occupied by the arrows. The closing force illustrated by the arrows has been referred to in the past as the ‘mechanical advantage’ of the subsidiary/epidermal cells (DeMichele & Sharpe 1973; Wu et al. 1985), or the ‘antagonism ratio’ between guard and subsidiary/epidermal cells (Cook et al. 1976). Due to the difficulties involved in its measurement there has been a lot of uncertainty about both the definition and the actual values of mechanical advantage m. Probably the most meaningful definition of the mechanical advantage is the partial derivative form given by Cook et al. (1976), which relates stomatal aperture a, guard cell pressure Pg and epidermal pressure Pep as follows:

Figure 7.

. Plots of T. virginiana stomatal aperture, a, as functions of guard cell hydrostatic pressure, Pga = f1(Pg) for epidermal turgor of zero. a = f2(Pg) for maximum epidermal turgor (in this case 0·92 MPa). Functions f1(Pg) and f2(Pg) were obtained by fitting sigmoidal curves to the data in Fig. 4. Details of these functions are given in Appendix 1.

inline image

We cannot estimate m precisely from our data which, as they relate to two states of epidermal pressure only, do not reveal the exact relationship between a and Pep. However, examination of Fig. 7 indicates that m varies widely with Pg, and probably with Pep also. Mechanical advantage, as defined above, is therefore difficult to visualize over the entire range of stomatal movement. Also, the present definition of mechanical advantage is difficult to interpret strictly in mechanical terms. This is because aperture has been chosen as a matter of convenience to represent displacement of the guard cell. While it is evident that guard cell displacement is occurring for all increments in Pg under all conditions, equating this displacement with changing aperture distorts the true nature of the mechanical process. A better approximation might be to use displacement of the dorsal wall instead of aperture in Eqn 1. Using dorsal wall displacement, the shape of the response with epidermal back pressure might perhaps resemble the shape of the response without. Be that as it may, stomatal aperture, being closely related to leaf diffusive conductance, is the more useful parameter in so far as one is primarily interested in the effect of guard cell movement on leaf gas exchange. It is then perhaps best to avoid formal use of the concept of mechanical advantage, and express observations of the kind we have made in the following way. Defining the dependences of aperture, a, on Pg with zero epidermal turgor and with maximum epidermal turgor as f1(Pg) and f2(Pg), respectively, and assuming reduction in a due to epidermal turgor pressure Pep is proportional to that pressure, it follows that stomatal aperture may be expressed as

inline image

where Pep(max) is maximum epidermal turgor. Examples of f1(Pg) and f2(Pg) for Tradescantia virginiana, as plotted in Fig. 7, are given in Appendix 1. With some knowledge of how epidermal and guard cell turgor are affected by leaf processes (e.g. transpirational fluxes), Eqn 2 could be very useful in mechanistic models of stomatal function.

The above results reveal new information about the mechanics of stomatal movement. They help to clarify previous uncertainties regarding the inflation of guard cells and the physical interaction between stomatal guard cells and epidermal cells. Different relationships between stomatal aperture and guard cell pressure may be a function not only of guard cell architecture, but also of the basic chemical constituents of guard cell walls. Similar experiments investigating the relationship between guard cell inflation characteristics and guard cell wall chemistry in different taxa could provide better insight into their stomatal limitations. In a broader sense these data may be useful in providing leaf gas exchange models with a mechanistic basis.


We would like to thank S. C. Wong, W. Coupland and P. Groeneveld for technical assistance. We are also grateful to Professor Brian Gunning for providing microscope facilities.


  1. Present address: Department of Tropical Plant Sciences, James Cook University, PO Box 6811, Cairns QLD 4870, Australia.



The technical graphics application ‘Origin’ (Microcal Software Inc., Northampton, USA) was used to perform a least squares fit of a standard sigmoidal function to the data. This function is of the form


y = ––––––––– + A1,

1 + e(x0–x)/dx


A1 = initial y value: y(–∞)

A2 = final y value: y(+∞)

x0 = centre

Dx = width

In this instance, y is stomatal aperture (a), and x is guard cell pressure (Pg).

For T. virginiana with zero epidermal turgor, A1 = –80·56, A2 = 19·43, x0 = –1·883 and dx = 1·385. At full epidermal turgor (Pep = 0·92 MPa), A1 = –0·440, A2 = 9·534, x0 = 2·551 and dx = 0·415.