Unusual stomatal behaviour on partial root excision in wheat seedlings



    1. Institute of Biology, Ufa Scientific Center, Russian Academy of Sciences, pr. Oktyabrya 69, 450054 Ufa, Russia and
    2. Plant Research Unit, Division of Environmental and Applied Biology, School of Life Sciences, University of Dundee at SCRI, Invergowrie, Dundee DD2 5DA, UK
    Search for more papers by this author

    1. Institute of Biology, Ufa Scientific Center, Russian Academy of Sciences, pr. Oktyabrya 69, 450054 Ufa, Russia and
    Search for more papers by this author

    1. Institute of Biology, Ufa Scientific Center, Russian Academy of Sciences, pr. Oktyabrya 69, 450054 Ufa, Russia and
    Search for more papers by this author
  • H. G. JONES

    Corresponding author
    1. Plant Research Unit, Division of Environmental and Applied Biology, School of Life Sciences, University of Dundee at SCRI, Invergowrie, Dundee DD2 5DA, UK
      Hamlyn G. Jones. Fax: + 44 (0)1382 568575; e-mail: h.g.jones@dundee.ac.uk
    Search for more papers by this author

Hamlyn G. Jones. Fax: + 44 (0)1382 568575; e-mail: h.g.jones@dundee.ac.uk


The excision of four out of five primary roots of wheat (Triticum durum Desf.) seedlings often leads to an enhanced rate of transpiration. Surprisingly this enhancement could be maintained for several hours after root excision and was particularly likely to occur at low irradiances or high atmospheric humidity. This long-term enhancement could not be explained in terms of conventional hydropassive stomatal effects. Elevated rates of transpiration were associated with and possibly caused by increased cytokinin concentrations in shoots of plants with partially excised roots. The single root remaining after excision was able to maintain an adequate water uptake for the continued enhanced transpiration, after only a short transient reduction in leaf water content. The enhanced capacity for water uptake by the remaining root was confirmed by measuring the water flow from detached roots at negative hydrostatic pressure. Even without additional suction, flow from the reduced root system increased about 1.5 h after the start of treatment, suggesting an increase in membrane permeability for water. Although abscisic acid (ABA) concentrations in the roots increased after the root excision treatment, there was no evidence for any enhanced concentration in the xylem sap. The possible role that this accumulation of ABA in roots may have in the apparent increase in hydraulic conductivity after root excision is discussed.


Maintenance of optimal tissue water content by plants requires a balanced control of both water uptake and water losses. Numerous environmental stresses met by plants during their ontogenesis tend to create water deficits; maintaining water relations under these conditions is important for plant survival. The ability of plants to close their stomata in response to developing water deficits is one of the main mechanisms enabling restoration of tissue water relations by limiting water losses under water deficit and this process has been intensively studied (see, e.g. Willmer 1983; Jones 1992; Kramer & Boyer 1995). There has been particular interest in recent years in the balance between hydraulic mechanisms involved in the control of stomatal aperture and the possibility that chemical signalling may play an important role (Davies & Zhang 1991; Davies, Wilkinson & Loveys 2002; Sperry et al. 2002). In general, reductions in water availability such as those that occur through soil drying or increased salinity tend to decrease the stomatal conductance to water vapour, thus tending to maintain tissue water status in a homeostatic manner. One might expect similar responses to reduced hydraulic conductance and to root removal (Meinzer 2002). Hydraulic signals have been particularly strongly implicated in the initial transient stomatal opening response that occurs when the water supply to a leaf is reduced; this effect was noted by Iwanov (1928) and has been replicated many times (e.g. Jones 1999) with the accepted explanation involving a hydropassive effect dependent on the subsidiary cells drying transiently before the guard cells (e.g. Meidner & Mansfield 1968; Raschke 1970). A characteristic of this response, however, is that any opening response is only transient.

Whatever the mechanism that is involved in the control of stomatal aperture in drought conditions, there is a clear trade-off between water conservation through stomatal closure and the consequent reduction in photosynthesis (e.g. Jones 1993). It is not surprising therefore that plants have developed a number of alternative mechanisms for the control of water relations such as increased allocation to root growth or changes in hydraulic conductivity. Growth responses, however, take time and may be inadequate in cases of the rather frequent sudden changes of environment (Munns et al. 2000). Changes in hydraulic conductivity in accordance with transpiration demand may provide an alternative mechanism for maintaining water relations (Weatherley 1982). Until recently it remained unclear how this might be achieved in plants, although several hypotheses have been proposed (Steudle & Peterson 1998) and the specialized water channels in plants (aquaporins, Schäffner 1998) may have a role. Nevertheless it still remains unclear how changes in hydraulic conductivity and transpiration are co-ordinated.

Root excision inevitably disturbs the balance between water uptake and loss and a study of the plant reaction to such a treatment should contribute to a better understanding of the homeostatic control of plant water relations. In our earlier experiments wheat seedlings have been shown to be able to maintain high rates of growth and transpiration and high water content after the removal of four seminal roots out of five (Vysotskaya et al. 2001). In these experiments, however, the cut root surface was placed in the nutrient solution, which opens an apoplastic bypass for water. It was considered of interest to follow the water relations of plants after partial root excision in an experiment designed to prevent penetration of water through the cut surface.

The present article arises from a chance observation made during a series of experiments on the water relations of wheat seedlings when part of the lighting system failed. In this case we observed a surprising increase in transpiration when some of the roots were severed, initially mimicking the Iwanov effect, but to our surprise the initial enhancement of transpiration was maintained for several hours. This study describes a series of experiments designed to confirm and further investigate the conditions giving rise to this maintained enhanced transpiration rate.


Seedlings of durum wheat (Triticum durum Desf., cv. Bezenchukskaya 139) were grown for 7 d in containers filled with 0.1 strength Hoagland–Arnon nutrient solution under an irradiance in the photosynthetically active region (PAR) of 400 µmol m−2 s−1 for a 14-h photoperiod and at 23–25 °C. All experiments were conducted at normal laboratory temperatures (23–25 °C) and humidities [relative humidity (RH) 35–45%– equivalent to vapour pressures of about 1.3 kPa]. Different transpirational demand was established by changing the air relative humidity and/or the illumination, while water uptake was also manipulated by the addition of mannitol to the nutrient medium. For this purpose measurements were carried out under constant illumination of between 50 and 500 µmol m−2 s−1, both during any experiment and for the previous 4 h acclimation period. At the beginning of each experimental run four seminal roots out of five were excised and the plants mounted so that the cut root surface was above the level of the nutrient solution or placed into it. Similar experiments were carried out both at Dundee and at the Ufa Research Centre with very similar results. Only a representative sample from many experiments (a total of 49 in Ufa and 34 in Dundee) is presented here.

Transpiration was measured continuously as the loss of water by one or five plants placed in container with 50 mL of nutrient solution measured by weighing every 2 min (the containers were covered with Parafilm to prevent evaporation from the water surface). Water content was measured as the difference between fresh and dry weight expressed as a percentage of fresh weight. Stomatal conductances were measured using a CIRAS 1 (PP Systems, Hitchin, Herts, UK). For following conductance changes after root excision, the leaves were maintained continuously in the leaf chamber and records taken every 2 min under different illumination (50–100 and 500 µmol m−2 s−1).

An analogue inductive electromechanical position sensor was used to monitor extension growth of the leaf as described earlier (Veselov et al. 2002). Output signals from the sensor were tracked continuously using a chart recorder.

For root exudate collection, the wheat seedlings were cut under water and the shoots reconnected to the roots with fine silicon tubes (Fig. 1). Typical transpiration flows of approximately 70 µL per seedling per hour, refilled the 10 µL void of the silicon tubing with xylem sap in about 8 min. A comparison of the transpiration rate of intact and cut and rejoined plants by the weighing technique showed that this procedure did not affect transpiration significantly. After 10 min of collecting root exudate the tubes from 10 plants were disconnected and the root exudate within the tubes weighed and sampled for determination of abscisic acid (ABA) concentration and cytokinin concentrations. Export of cytokinins to the shoot was calculated by multiplying the measured cytokinin concentration in root exudate by the transpiration rate.

Figure 1.

Scheme of xylem sap collection: 1, position of aerial part excision; 2, position of root excision; 3, silicon tube; 4, position of a seedling in nutrient solution.

For hormone extraction, the roots were homogenized in 80% ethanol and incubated over night at +4 °C. After filtration and vacuum evaporation of extracts to remove all traces of ethanol the aqueous residue was divided and one part further processed for cytokinin purification as described by Vysotskaya et al. (2001). Cytokinins from the aqueous residue were concentrated on a pre-wetted C18 column (Bond-Elut®, RP-C18; Varian Ltd., Walton-on-Thames, UK) and, after washing, the column was loaded with 20 mL of distilled water and then eluted with 5 mL of 80% ethanol. After solvent evaporation, the dry residue was dissolved in 0.02 mL of 80% ethanol and applied to pre-coated Merck 50 × 200 × 0.25 mm silica gel 60 F-254 thin-layer chromatography (TLC) plates [Merck (UK) Ltd., Poole, Dorset, UK] which were developed in 2-butan-ol: 14 m NH4OH: H2O (6 : 1 : 2 v/v, upper phase). Cytokinin-containing zones (based on position of standards) were eluted with 0.1 m phosphate buffer, pH 7.4 for 12 h and added directly to microplate wells in serial dilutions. These were assayed using antibodies against zeatin riboside (ZR), which have been shown to be highly specific to several zeatin derivatives (Kudoyarova et al. 1998). This protocol successfully separated and assayed zeatin nucleotide (Rf 0–0.1), zeatin riboside (Rf 0.4–0.5), and zeatin (Rf 0.6–0.7). More than 90% recovery was obtained for zeatin, its riboside and glucoside standards using the described elution procedure. The sum of the listed cytokinins is presented.

For the ABA assay, aqueous residue was diluted with distilled water, acidified with HCl to pH 2.5 and was partitioned twice with peroxide-free diethyl ether (ratio of organic to aqueous phases was 1 : 3). Subsequently the hormones were transferred from the organic phase into 1% sodium bicarbonate (pH 7–8, ratio of organic to aqueous phases was 3 : 1), re-extracted with diethyl ether, methylated with diazomethane and immunoassayed using antibodies to ABA (Veselov et al. 1992). ABA recovery calculated in model experiments was about 80%. Reducing the amount of extractant, based on the calculated distribution of ABA in organic solvents, increased the selectivity of hormone recovery and the reliability of the immunoassay.

The reliability of the immunoassay for ABA was ensured by both the high specificity of antibodies and pre-purification of hormones according to a modified scheme of solvent partitioning (Veselov et al. 1992). The anticytokinin antibodies had high immunoreactivity towards zeatin, its riboside, 9N-glucoside and nucleotides, which were separated by TLC prior to immunoassay (Veselov et al. 1999). The antibody showed low cross reactivity to dihydrozeatin and isopentenyladenine (iPA) and their derivatives. The reliability of all the hormone immunoassays was confirmed by dilution tests, chromatographic examination of the distribution of immunoreactivity and comparison of the results of immunoassay against that of physicochemical assays [liquid chromatography–mass spectometry (LC–MS), Veselov et al. 1999].

Samples for measurement of osmotic potential were obtained after freezing and thawing of roots and expressing sap by squeezing in a syringe. They were measured by means of vapour pressure osmometer (Model 5100C; Wescor Inc., Logan, UT, USA).

To measure the water flow from the detached root system, the aerial parts of the plant were removed under water leaving a cylinder of leaf bases, which were connected to a thin empty capillary by means of a silicon tube (the weight of each capillary connected to the silicon tube was measured prior to connecting to the root system). After 1 h the root system was disconnected and the weight of the capillary plus root exudate measured. Sap flow was expressed in m3 m−2 s−1 (Jackson 1993). Aerial parts were removed and capillaries attached, either immediately after root excision or 1.5 h later. Xylem sap flow was also measured under suction produced by vacuum as described by Freundl, Steudle & Hartung (1998). The excised root systems with the mesocotyl still attached were fixed to a capillary using a pressure-tight silicone seal fixed by a screw. Silicone seals were individually prepared for each plant. They had a length of 15 mm and a hole individually adapted to the diameter of the mesocotyl. The root medium was aerated. Suction was applied to the mesocotyl; this caused xylem sap flow into a calibrated capillary. Xylem sap exuded into the capillary was collected, weighed and analysed for hormone content after 15 min at −0.04 MPa. For all these experiments the cut root surface was placed above the nutrient medium.


Figure 2b shows only a slight (but significant) transitory decline in water content in the differentiated part of the leaf after roots were excised under a low illumination. No such effect was detected at the higher irradiance treatment (Fig. 2d). In the growing leaf part the decline in water content was more pronounced, being reduced by 2–2.5% 20 min after the treatment, and remaining lower than in the control throughout the subsequent 1.5 h (Fig. 2a & c).

Figure 2.

Water content (% of fresh weight) in growing (a, c) and fully expanded leaf zones (b, d) in intact plants and in those after partial root excision grown for 4 h prior to, and during, the experiment at 50–100 µmol m−2 s−1 PAR, RH 60–70% (a, b) or 400–500 µmol m−2 s−1 PAR, RH 45–50% (c, d) (means of 20 replicates with standard errors).

Typical dynamics of leaf growth responses to partial root excision under different environmental conditions are shown in Fig. 3. Cessation of leaf growth and some leaf shrinkage were observed immediately following root excision with the cut surface placed above nutrient solution (Fig. 3a & b). Where, however, the cut root surface was placed inside nutrient solution (Fig. 3c), no cessation of leaf growth was recorded and leaf growth was actually enhanced.

Figure 3.

Representative traces showing the effect of partial root excision (indicated by the arrows) on leaf growth rate of wheat plants grown for 4 h prior to, and during, the experiment at 400 µmol m−2 s−1 (a, c) or 100 µmol m−2 s−1 (b) with cut surface placed above (a,b) or inside (c) the nutrient solution.

The results from a large number of measurements of transpiration are shown in Fig. 4. The effects on stomatal conductance are summarized in Fig. 5, with some examples of actual curves shown in Fig. 6. It is notable that a subjective classification of the responses showed significant variation in the responses observed, although the general difference with evaporative demand was clear. In almost all the cases an initial very fast increase in transpiration and stomatal conductance was detected, which was often maintained for periods of an hour or more. Of particular note is the observation that the increase in transpiration or conductance was greater, and maintained for longer under low irradiance conditions and at higher humidity. Indeed, at high humidity the enhancement continued for the full two or more hours of each experiment in most cases.

Figure 4.

Effect of partial root excision (time of treatment indicated by arrow) on transpiration of wheat seedlings grown for 4 h prior to, and during, the experiment at 50–100 µmol m−2 s−1 PAR (a) or 400 µmol m−2 s−1 PAR (b) and different air humidities as indicated in the figure legends (means of 10 (a) and 30 (b) replicates ± standard errors).

Figure 5.

Scheme showing a subjective classification of the various types of stomatal conductance (Gs) response to root excision depending on transpiration demand. The number of plants responding in each of the characteristic patterns is indicated on the figure legend. (a) Low irradiance and high humidity; (b) low irradiance and low humidity; (c) high irradiance and high humidity; (d) high irradiance and low humidity.

Figure 6.

Some typical examples of the effect of partial root excision (time of treatment indicated by arrows) on stomatal conductance of wheat seedlings. During measurements whole plants were under laboratory conditions (20 µmol m−2 s−1 PAR), the measured parts of leaves were maintained continuously in the leaf chamber and records taken every 2 min at 50–100 µmol m−2 s−1 PAR (a, b) or 500 µmol m−2 s−1 PAR (c, d) with 35% (a, c) or 70% (b, d) relative air humidity.

Figure 7 shows that lowering the water potential of the nutrient medium changed the reaction of plants to partial root excision. In this case, after an initial short transient, there was a consistent stomatal closure in response to partial root excision.

Figure 7.

Some examples of effect of adding mannitol to produce 5% concentration (= 0.70) MPa (first arrow) and partial root excision (second arrow) on stomatal conductance (Gs) in wheat seedlings. During measurements whole plants were under laboratory conditions (20 µmol m−2 s−1 PAR) with the studied parts of leaves maintained continuously in the leaf chamber and records taken every 2 min at 500 µmol m−2 s−1 PAR.

Data on changes in xylem sap ABA concentrations are presented in Table 1. ABA concentration in xylem sap was reduced by partial root excision under low illumination but remained similar to the control under high illumination (Table 1). An increase in ABA content in root systems was detected some hours after partial root excision (Fig. 8), with the maximum increase occurring around 2 h after treatment. The changes in shoot cytokinin content (Fig. 9) clearly demonstrated an increased cytokinin concentration in the shoot tissue after root excision, and that this difference was also maintained for several hours. An initial decline in cytokinin concentration and rate of delivery was observed in the xylem sap 1 h after partial root excision (Table 2). A day after treatment both hormone concentration and delivery rate had returned to control values.

Table 1.  Concentration of ABA in xylem sap (nm) collected during 10-min periods before root excision, and 40 min after root excision for plants grown for the 4 h prior to, and during, the experiment at irradiances of 50–100 or 400 µmol m−2 s−1 PAR
 100 µmol m−2 s−1 PAR, 60–70% RH400 µmol m−2 s−1 PAR 45% RH
  1. Means (n = 9) of three sets of experiments of three replicates each with 10 plants sampled for each measuremen,t together with standard errors. Control samples were collected from intact plants simultaneously with samples obtained from excised plants.

Before root excision109.7 ± 19.628.0 ± 4.9
40 min after root excision120.7 ± 18.257.1 ± 13.740.9 ± 11.042.0 ± 11.7
Figure 8.

Effect of partial root excision on root ABA content of wheat plants grown 4 h prior to experiment and measured at irradiances indicated in the figure legends. Results are means ± standard errors (n = 9) for three sets of experiments with three replicates, each with 10 plants sampled for each measurement.

Figure 9.

Effect of partial root excision on cytokinin content in shoots of plants grown 4 h prior to experiment and measured at irradiances indicated on the figure legends. Results are means ± standard errors (n = 9) for three sets of experiments with three replicates, each with 10 plants sampled for each measurement.

Table 2.  Concentration of cytokinins in xylem sap and their delivery rate from roots of intact and treated plants grown for 4 h prior to, and during, the experiment at irradiances of 50–100 and 400 µmol m−2 s−1 PAR
Concentration (nm)Low light (100 µmol m−2 s−1)High light (400 µmol m−2 s−1)
1 h after root excision52.5 ± 14.8 9.9 ± 3.0385.8 ± 35.7 86.3 ± 17.9
24 h after root excision74.4 ± 17.969.5 ± 14.9342.3 ± 32.7356.1 ± 38.7
Delivery rate
[pmol s−1 ( × 10−5)]
Low light (100 µmol m−2 s−1)High light (400 µmol m−2 s−1)
  1. Means (n = 9) of three sets of experiments three replicates each with eight plants sampled for each measurement, together with standard errors. Control samples were collected from intact plants simultaneously with samples obtained from excised plants.

1 h after root excision16 ± 47 ± 1769 ± 99239 ± 49
24 h after root excision33 ± 526 ± 4761 ± 91612 ± 107

Rates of water flow from the root system under vacuum (Fig. 3) were used to investigate the changes in hydraulic conductivity of the whole root system (Table 3). Immediately after root excision water flow was halved in normal plants, but when placed under vacuum, water flow increased at least four-fold in comparison with the intact plant also put under vacuum. In the absence of negative pressure, water flow was faster from the reduced root system than from the intact system at 1.5 h after treatment.

Table 3.  Effect of partial root excision on water flow from roots (m3 m−2 s−1) of plants grown at irradiances 400 µmol m−2 s−1 PAR
 Negative pressure applied (MPa)
0− 0.04
  1. Means of nine replicates ± standard errors. Shoots were removed under water leaving a cylinder of leaf bases, which were then connected to empty tubes for root sap collection naturally or under vacuum.

Intact root system 7.5 ± 1.7 ( × 10−10)18.8 ± 2.7 ( × 10−10)
5 min after root
 3.2 ± 1.1( × 10−10)77.6 ± 10.0 ( × 10−10)
1.5 h after root
30.1 ± 4.2 ( × 10−10)91.7 ± 10.1 ( × 10−10)

The osmotic potential of the roots did not change following partial root excision [215 ± 9 mosmol kg−1 (n = 8) in intact plants and 220 ± 11 (n = 9) in those with partially excised roots]. The xylem sap pH 20 min after root excision increased from 6.9 ± 0.1 (n = 5) in control plants to 7.3 ± 0.1 (n = 5) in treated plants.


Conventional understanding of plant water relations led us to expect that partial root excision would decrease the water uptake capacity of the plants, which in turn should result in increased shoot water deficit, and hence stomatal closure. In order to test these expectations we first measured changes in water content following root excision in plants grown under conditions of high (400–500 µmol m−2 s−1 PAR; RH, 45–50%) and low (50–100 µmol m−2 s−1 PAR; RH 60–70%) transpiration demand.

Leaf water content was decreased by partial root excision, with the effect being more pronounced in the growing than in the mature leaf tissue (Fig. 2). A similar greater sensitivity of growing tissue as compared with mature tissue has been noted previously (Matsuda & Riazi 1981). A further indicator of increased water deficit can be obtained from fast changes in leaf growth (Chazen & Neuman 1994; Munns et al. 2000), since cell expansion depends on water uptake. The results suggest that plants with reduced root systems started to experience water deficit even quicker than the measurable decline in water content. The assumption that the cessation of leaf growth resulted from the increased leaf water deficit is supported by the experiment (Fig. 3c; see also Vysotskaya et al. 2001), where the cut root surface was placed inside nutrient solution. In this case no water deficit was observed, since water penetrated directly into the cut xylem vessels, and leaf growth was not inhibited but actually increased (Fig. 3).

The changes in growth and water content were small and largely disappeared within 1.5 h after partial root excision. It was therefore of particular interest to understand how plants with only one remaining root (one fifth of the normal complement) managed to maintain their water status. Changes in stomatal conductance provide the most important short-term control of water relations when the balance between water uptake and loss is disturbed (Meidner & Mansfield 1968; Willmer 1983) in plants following partial root excision. Although the pattern of changes in transpiration and stomatal conductance at different illumination and humidity was somewhat variable (Figs 4, 5 & 6), quite unexpectedly a decline in stomatal conductance was almost never seen and a maintained increase in transpiration and stomatal conductance were frequently observed instead.

What might have caused the increase in transpiration? The initial leaf shrinkage observed following partial root excision, reflects a decline in water delivery from the reduced root system and might cause hydropassive stomatal opening. The usual explanation of this phenomenon (e.g. Raschke 1970) is that a decrease in epidermal and subsidiary cell turgor releases back pressure on the guard cells, allowing stomatal opening and enhanced transpiration. It seems unlikely that this localized water imbalance could account for any maintenance of elevated stomatal conductance beyond a few minutes (e.g. Sharpe, Wu & Spence 1987; Saliendra, Sperry & Comstock 1995). When water deficits and leaf shrinkage are prevented (Fig. 3 and Vysotskaya et al. 2001) by leaving the cut root surfaces under solution no maintained stomatal opening was observed. Therefore the present results showing that transpiration and stomatal conductance do not in general fall below the values observed before root excision, and under some circumstances are maintained at higher values for periods of several hours or more (e.g. Figs 4, 5 & 6), especially under conditions of low transpirational demand, require some alternative explanation.

The hormone analyses provide some important indications as to what might be happening. Root-sourced ABA has long been implicated in stomatal responses to soil drying (Davies & Zhang 1991). ABA loading into xylem of one root was seemingly increased by partial root excision, which is indicated by the fact that the concentration of ABA was reduced only two-fold whereas the number of roots was one-fifth that in intact plants. Nevertheless, a reduction in the size of the hormone-producing organ is at least partly responsible for the observed decline in ABA concentration in the xylem sap, which in turn might contribute to maintaining stomata open under low illumination. With high illumination there was no decline in ABA concentration in xylem sap nor was the elevated transpiration maintained.

The pH of xylem sap is also known to influence stomatal conductance and transpiration (Hartung 1983; Hartung, Radin & Hendrix 1988; Slovik & Hartung 1992; Wilkinson 1999) but the observed increase in pH would be expected to close rather than open stomata.

On the other hand, the observed increase in cytokinin content (Fig. 9) of shoot tissues is in the direction required to explain increased transpiration (Blackman & Davies 1985; Mansfield & McAinsh 1995). However it is unclear how that enhanced shoot cytokinin arose, as root excision removes cytokinin-producing tissue, and cytokinin concentration in xylem sap (and delivery rate to the shoot) were reduced 1 h after partial root excision (Table 2) in proportion to root loss. It is possible that the shoot concentration was maintained through changes in cytokinin metabolism in shoots. This suggestion is supported by evidence that the activity of cytokininoxidase can be inhibited by partial root excision (Vysotskaya et al. 2001); this change might be responsible for the cytokinin accumulation (Fig. 9). In the longer term, elevated cytokinin concentrations (Table 2) may be maintained through an increased ability of the remaining roots to produce cytokinins because of the initiation of numerous lateral roots observed very shortly after partial root excision (Vysotskaya et al. 2001).

Although an early increase in transpiration or stomatal conductance was common in all conditions, maintenance of elevated transpiration for extended periods was characteristic of low evaporative demand environments (high humidity or low irradiance). Continued maintenance or enhancement of transpiration after root removal requires an increased capacity for water uptake by the remaining roots; only under low evaporative demand was this adequate.

The enhanced transport capacity of roots does not appear to result from a change in driving force as there was no evidence for any effect on the overall osmotic potential of roots following partial root excision. Consequently, at least the increase in water flow in detopped systems in the absence of applied negative pressure 90 min after partial root excision is likely to be due to increased hydraulic conductivity. In such a system the increased hydraulic conductivity probably reflects changes in permeability of membranes for water (Steudle & Peterson 1998), which may reflect altered aquaporin activity through changes in phosphorylation state (Schäffner 1998), although an increase in the contribution of the apoplastic pathway to water transport may also be involved (Steudle & Peterson 1998). Since partial root excision should increase xylem tension in the only remaining root, this might be the cause of the initial increase in hydraulic conductivity, while subsequent increases in membrane permeability might also contribute to maintaining its high level.

The increased ABA concentrations observed after root excision may have contributed to the increased hydraulic conductance (Fiscus 1981; Freundl et al. 1998). It seems unlikely, however, that the initial increase in hydraulic conductivity depended on ABA, as the observed increase in ABA content in root systems was only apparent some hours after partial root excision (Fig. 8).

In order to confirm the importance of water uptake in maintaining water relations in plants with a partially reduced root system, we also investigated the effect of alterations to the osmotic gradient by addition of mannitol to the nutrient medium (Fig. 7). In this case the sole remaining root was not able to maintain the water flow required, even though the experiment was conducted under otherwise optimal conditions for the effect to be apparent (high relative humidity).

Although the present experiments were carried out on solution-grown plants, there are several reports on the ability of soil-grown plants to maintain their growth despite reductions in their root system (e.g. Shane & McCully 1999). Our results provide clear evidence that variation in root hydraulic conductance can provide a powerful mechanism to compensate for limitations in water supply allowing plants to maintain an adequate water balance; these mechanisms complement stomatal control of water loss. The increase in root conductance in accordance with transpirational demand is an alternative mechanism for the control of water balance with the advantage that it allows stomata to remain open permitting continued gas exchange and photosynthesis.

Triticum durum plants are known to be more drought tolerant than other wheat species, so we are not able to predict whether the present results would be applicable to other more mesophytic wheat species.


We are grateful to the Royal Society for funding for an Ex-Agreement Visit to the UK for L.V. This work was supported by a grant of Russian Foundation of Basic Researches N 03-04-49780.