Influence of pressurized ventilation on performance of an emergent macrophyte (Phragmites australis)

Authors


Viveka Vretare (fax + 46 462224536; e-mail viveka.vretare@limnol.lu.se).

Summary

1 Pressurized ventilation, which increases gas exchange between aerial and submerged plant parts, has been found in various emergent macrophyte species. We investigated the potential for this mechanism to affect growth, morphology and biomass allocation in Phragmites australis in glasshouse experiments.

2 Inhibition of pressurized ventilation by perforation of stems above the water surface resulted in decreased oxygen concentrations in stem bases and rhizomes. Perforation caused little mechanical damage.

3 Allometric methods were used to evaluate treatment effects on biomass allocation and morphology.

4 Inhibition of pressurized ventilation resulted in decreased allocation to below-ground weight and decreased rhizome penetration into the substrate in two of three experiments. Treatment also decreased growth rate, rhizome length and number of rhizomes when substrate had a high organic content. In the third experiment, growth clearly decreased in deep water, although inhibition of pressurized ventilation did not affect growth, biomass allocation or morphology at either of the water depths tested.

5 Decreased allocation to below-ground parts and decreased rhizome lengths may be adaptations to allow the oxygen concentration in roots and rhizomes to be maintained above a critical level when the oxygen supply is low.

6 Pressurized ventilation may improve the performance of P. australis but only under certain conditions (e.g. not when growth rate is low or the substrate has a high redox potential).

Introduction

Wetlands and littoral zones are often dominated by large, rhizomatous emergent macrophytes. These plants generally have their below-ground tissues in anoxic sediments, and their roots are thus dependent on oxygen transport from the aerial parts, to enable aerobic respiration (Drew et al. 1985) and efficient nutrient uptake (Bradley & Morris 1990; Koch et al. 1990), as well as for protection against phytotoxins (Engler & Patrick 1975; Begg et al. 1994; Wang & Peverly 1999) as these form readily in sediments with low redox potential (Laanbrock 1990; Cizkova et al. 1999). Oxygen deficiency in below-ground parts may limit the depth penetration of Zizania latifolia (Griseb.) Stapf. (Yamasaki 1984), and the growth rate of rice (Oryza sativa L.) roots has been shown to decrease when the oxygen supply is low (Armstrong & Webb 1985). Furthermore, the growth rate of emergent macrophytes has been shown to decrease in substrates with low redox potential compared to well aerated conditions (Kludze et al. 1993; Brix & Sorrell 1996; Weisner 1996).

In some species of emergent macrophytes, bulk flow of oxygen occurs between plant parts differing in total pressure of the internal gas space (Armstrong & Armstrong 1991; Brix et al. 1992; Sorrell & Tanner 2000). This pressurized ventilation is supplementary to, and may potentially be much more effective than, diffusion as a mechanism supplying the below-ground tissues with oxygen (Armstrong & Armstrong 1990; Sorrell et al. 1997). Because pressurized ventilation is found in many different genera, including Typha (Bendix et al. 1994), Phragmites (Armstrong & Armstrong 1991), Eleocharis (Brix et al. 1992), Cyperus (Brix et al. 1992) and Spartina (Hwang & Morris 1991), it could be a common adaptation to anoxic sediments (Brix et al. 1992). The direction of flow can be from green stems, via the rhizomes, either to dead or damaged stems or to other stems with lower pressurizing ability (Armstrong et al. 1996b). New rhizomes (which do not yet have stems emerging above the water surface) and roots do not provide an outlet through which gas can escape, and therefore no through-flow ventilation can occur through these parts. Consequently, apices of these parts must acquire oxygen mainly by diffusion, although this is possibly enhanced by the non-throughflow ventilation reported by Raskin & Kende (1985) and Frick et al. (1997). In a review of the importance of O2 in soils, Drew (1990) points out that the actively growing apices of roots and rhizomes are likely to be particularly sensitive to low O2 concentrations. Although they cannot take direct advantage of pressurized ventilation, to ensure an adequate supply, there may be an indirect effect as the enhanced O2 concentration in old rhizomes lead to an increased concentration gradient from the rhizome to the root apex, or through a new rhizome, and thus to increased diffusion of O2 to roots and new rhizome apices (Armstrong & Armstrong 1990, 1991).

Although the mechanisms and occurrence of pressurized ventilation in emergent macrophytes have been well investigated (see, for example, Armstrong & Armstrong 1991; Armstrong et al. 1992; Brix et al. 1992; Bendix et al. 1994), little is known about its importance for plant performance (e.g. growth rate, dispersal rate, ability to grow in deep water). The aim of this study was to investigate whether pressurized ventilation affects the performance of Phragmites australis (Cav.) Trin. ex Steud., a cosmopolitan species which is common in temperate wetlands all over the world (van der Toorn 1972) and is known for its high rates of pressurized ventilation (Brix et al. 1992).

Methods

Pressurized ventilation was inhibited by using an injection needle to make holes (1 mm wide and 3–6 mm long) connecting the pith cavity of the stem with the ambient air. Three experiments were conducted to investigate the influence of pressurized ventilation on plants growing in a high-redox substrate (experiment I), at two different water depths (experiment II) and in a low-redox substrate with an increased need for O2 transport (experiment III). The effects of damage due to perforation were assessed in experiment I, and cutting of one stem just above water line (experiment II) or further perforation (experiment III) was used in some controls to ensure efficient throughflow ventilation.

All experiments were performed in a greenhouse with a mix of natural light and artificial light (Vialox Planta-T, 500 W). A 16-h photoperiod (0600–2200) was imposed and when outdoor light intensity went below 250 µmol s−1 m−2 PAR during this time, the artificial light was automatically switched on (to give an additional 150–200 µmol s−1 m−2 PAR). The daily mean temperature in the greenhouse during the experiment was between 21°C and 24 °C (min. 15 °C, max. 33 °C) and the mean relative humidity was 62% (min 24%, max 76%). Phragmites australis plants were raised from seed in wet/flooded potting soil and replanted in vermiculite and given an application of commercial fertiliser (20% N; 2.2% P; 6.6% K; 1.2% Mg; 0.5% micro nutrients, Super Gramino®, Weibulls, Sweden) before initiation of the experiments.

Experiment i

Three 7-week-old plants (initial dry weight (DW) = 0.91 ± 0.20 g (mean ± SD, n = 5)) were planted in each of 24 containers (diameter 70 cm, height 100 cm) filled with vermiculite supplemented with 50 g fertiliser, to a depth of 20 cm. The water level was kept at the substrate surface until the plants were 9 weeks old after which the water level was raised over a period of 2 weeks to the maximum depth of 80 cm such that all plants had at least one stem with one entire leaf sheath above the water surface during the entire experiment. When the plants were 12 weeks old, containers were allocated to one of three treatments. Stems were either all perforated close to the water surface (8 containers, Fig. 1a), perforated but the holes were then sealed with grafting-wax (Scheidler LacBalsam®) (8 containers) or served as non-perforated controls (8 containers, Fig. 1b). Treatments were reapplied to stems newly emerging above the water surface on every third day until plants were 16 weeks old and the experiment was terminated. The number of stems, adventitious stems and new rhizomes (those with stems that had not emerged above the water surface) were counted. For new rhizomes, the depth in the sediment, diameter at the middle of the longest internode and horizontal length were measured. Roots, shoots, rhizomes and aquatic roots were dried (105 °C, 48 h) separately and combusted (450 °C, 2 h) to determine ash free dry weight (AFDW). As it was difficult to remove all vermiculite from the roots in this experiment, AFDW was used for further analyses instead of DW.

Figure 1.

Schematic diagram to show which submerged parts of P. australis may have experienced gas exchange mediated by pressurized ventilation (white areas) rather than mainly diffusion (black areas) in each of the three glasshouse experiments. All treated stems were perforated (indicated by black dots) and all gas exchange therefore occurred by diffusion (a), whereas in the controls (b–d) pressurized ventilation occurred through stems and old rhizomes but not through roots and terminal rhizomes. When all stems were intact (b) pressurized ventilation could occur between plant parts of differing pressurizing ability (Experiment I, some controls in Experiment II). Creation of gas outlets (arrows) by cutting one stem per plant above the water surface (c) (some controls in Experiment II) or perforating 20% of the stems (d) (all controls in Experiment III) was used to facilitate pressurized ventilation.

Experiment ii

Nine-week-old plants with an initial DW of 6.0 ± 1.2 g (mean ± SD, n = 11), were planted individually in 72 buckets (10 L, height 26 cm, upper diameter 26 cm, basal diameter 21 cm), containing 25 cm depth of vermiculite and 5 g fertiliser. A two-way design was used with each of three main treatments imposed on 12 replicates of both perforated and control plants. Buckets were randomly distributed between 21 containers (as in experiment I) filled with aerated tap water with three or four buckets in each container. The buckets were attached by ropes to wooden frames, lying on top of the containers, and their water depth could therefore be adjusted independently. The substrate surface of replicates of the shallow water treatment was kept 5–7 cm below the water surface. Deep water or deep water with one stem per plant cut above the water surface (Fig. 1c) to facilitate outflow for the gas (simulating dead or broken stems) was simulated by submergence so that a minimum of two stems had at least 3 cm of leaf sheath above the water surface up to a maximum depth of 80 cm. Stems were perforated in the middle of the lowermost internode that emerged completely above the water surface, repeated if the hole became submerged, and new stems that had emerged above the water surface were perforated each day.

The experiment was terminated when the plants were 15 weeks old by which time the plants had unfortunately become infested by aphids. A 0.5-mL gas sample was gently withdrawn from pith cavities of internodes of horizontal rhizomes situated between emergent stems using a 5-mL syringe. Removal under the water surface avoided contamination of the sample by air. This sample was then injected via a narrow tube into a Teflon® chamber (0.3 mL) containing a Clark type electrode, calibrated with air and N2. The number of stems and adventitious stems were counted and the total length of the rhizomes were measured as well as the basal stem diameters. The DW of roots, shoots, rhizomes and aquatic roots were determined separately (105 °C, 48 h). In the treatment with deep water and one stem per plant cut above the water surface, the removed stem part was not included in the shoot weight. One plant (control, shallow water) which died during the first week of the experiment was excluded from further analysis.

Experiment iii

Twelve-week-old plants with an initial DW of 24.4 ± 4.2 g (mean ± SD, n = 10) were planted individually in 24 containers (as in experiment I and II) filled to 30 cm depth with vermiculite mixed with organic lake sediment (15% of the volume) and 10 g fertiliser. The water level was increased by 10 cm per day to a depth of 70 cm when treatments were initiated. For 12 plants, stems were perforated at the lowermost internode completely emerging above the water surface (Fig. 1a) and new stems were perforated when elongation led to parts not covered by leaf sheaths being exposed above the water surface. The remaining 12 control plants had 20% of their stems perforated (Fig. 1d). The perforated stems in the control plants simulated the dead or broken stems occurring in natural reedstands, thus enhancing the rate of pressurized ventilation by creating outlets for the gas. Perforation of new stems was performed daily. Redox potential in the substrate was measured at the beginning and end of the experiment (with a platinum electrode against a calomel reference electrode, 10 cm sediment depth, 5 min to equilibrate). The reference potential (+ 247.9 mV) was added to the measured value.

The plants were harvested after 3 weeks. A 0.5-mL gas sample was withdrawn from the lowermost internode above the sediment surface of five stems per plant and analysed as in experiment II. The stems from which gas samples had been taken were then cut at the water surface and the rate of outflow determined by connecting the portion removed to the upper vertical arm of a T-shaped glass tube with a cm-scale on the horizontal arm (length 220 mm, inner diameter 4.8 mm). The cut stem was kept upright, while the horizontal arm was filled with water using a pipette filler connected to the lower vertical arm; the flow rate was determined by measuring the rate at which the horizontal water column moved.

The number of stems and adventitious stems were counted, and the stem height from the substrate surface to the end of the uppermost leaf sheath and the basal stem diameters were measured. The number of new rhizomes and their horizontal length, depth in the substrate and diameter at the middle of the longest internode were determined. The DW of roots, shoots, rhizomes and aquatic roots were determined separately (105 °C, 48 h).

Statistical analyses

All plant biomass and morphological data were log(x) transformed to meet assumptions of normality and/or homoscedasticity except for number of aquatic roots and adventitious stems, which were log(1 + x) transformed. Oxygen concentrations and flow rates were not transformed. Normality was tested with Kolmogorov-Smirnov/Liljefors tests and homoscedasticity was tested with Levenes test. Analysis of variance (anova) was used to test for differences between treatments in total DW and O2 concentrations in stem bases and rhizomes. Allometric methods were used to handle the effects of differences in final plant size on allocation patterns (Coleman et al. 1994; Gedrock et al. 1996). Analysis of covariance (ancova) was used to test for differences between treatments in biomass allocation, rhizome architecture and number and diameter of stems. Log(total AFDW) or log(above-ground AFDW) were used as covariates in experiment I, and log(total DW) or log(above-ground DW) were used as covariates in experiments II and III. Non-significant (P > 0.05) terms involving covariates were dropped using backward elimination until the model was as simple as possible, improving the models power and interpretability. When interaction terms involving covariates were significant, a t-test was performed for least squares means (LSM) to determine if treatments differed significantly within the given range of the covariate. In experiment I, pair-wise two-tailed t-tests between LSMs were used to test which treatments that were significantly different from each other. The statistical package systat (systat Inc., Evanston, IL, USA 1992) was used for the Kolmogorov-Smirnov/Liljefors test and Levenes test, and Superanova (Abacus Concepts Inc. Berkeley, CA, USA 1989) was used for anovas, ancovas and t-tests between LSMs.

Results

Experiment i

There was no significant difference in total AFDW between treatments (one-way anova; F2,21 = 0.86, P > 0.05, Table 1). Significant effects of perforation were found in below-ground AFDW, rhizome depth and number of adventitious stems (Table 2). The interaction term between the treatments and total AFDW was significant when rhizome AFDW was the dependent variable. Comparisons between LSMs showed that within the given range of the covariate, perforated plants allocated less to below-ground (Fig. 2a) and rhizome biomass (Fig. 2b) and had more superficial rhizomes (Fig. 2c) compared to control and sealed plants. The LSMs for number of adventitious stems were significantly higher in the perforated and sealed treatments compared to the controls (Fig. 2d). No significant effects of the treatments were found in allocation to root AFDW, or aquatic root AFDW or in rhizome length, rhizome diameter, number of rhizomes or number of stems (Table 2).

Table 1.  Mean ± SD of variables measured in experiment I. The treatments were perforated (all stems were perforated to inhibit pressurized ventilation), sealed (all stems were perforated and the holes were sealed with grafting wax) and control (all stems were intact)
 ControlSealedPerforated
Total AFDW (g)220.1 ± 36.1196.8 ± 43.2205.5 ± 27.2
Rhizome AFDW (g)21.7 ± 7.823.0 ± 3.014.6 ± 2.4
Shoot AFDW (g)151.8 ± 22.0139.5 ± 35.7152.4 ± 23.2
Root AFDW (g)27.9 ± 5.521.2 ± 8.119.9 ± 6.9
Aquatic root AFDW (g)18.6 ± 9.613.1 ± 6.018.6 ± 7.7
Rhizome depth (cm)11.3 ± 1.610.9 ± 1.59.3 ± 1.6
Rhizome length (cm)12.4 ± 3.311.6 ± 3.79.8 ± 2.2
Rhizome diameter (mm)5.8 ± 0.55.5 ± 0.65.5 ± 0.4
Number of rhizomes26.0 ± 7.024.0 ± 3.724.4 ± 6.7
Number of stems110.8 ± 11.2110.5 ± 13.9113.6 ± 14.2
Number of adventitious stems1.9 ± 2.610.3 ± 6.315.0 ± 7.6
Table 2.  One-way ancovas for the effects of the treatments (perforated, sealed or control) in experiment I on different variables (dependent variables) with log(above-ground AFDW) or log(total AFDW) as covariates. The interaction term between the dependent variable and the covariate (D × C) is eliminated when not significant (NS, P > 0.05) to increase the power and interpretability of the test. The dependent variables are log(x) transformed, except for aquatic roots and number of adventitious stems, which are log(1 + x) transformed. *  P < 0.05, **  P < 0.01, ***  P < 0.001
Dependent Variable (D)Covariate (C)TreatmentD × C
  • a

    Term excluded as the interaction term (D × C) is significant

Above-ground AFDW  
Below-ground AFDWF1,20 = 2.20F2,20 = 7.23**NS
Total AFDW  
Rhizome AFDWF1,20 = 4.41__aF2,20 = 4.06*
Root AFDWF1,20 = 6.36*F2,20 = 2.53NS
Aquatic root AFDWF1,20 = 21.1***F2,20 = 1.01NS
Rhizome depthF1,20 = 1.87F2,20 = 3.52*NS
Rhizome lengthF1,20 = 0.55F2,20 = 0.79NS
Rhizome diameterF1,20 = 5.40*F2,20 = 0.38NS
Number of rhizomesF1,20 = 0.02F2,20 = 0.17NS
Number of stemsF1,20 = 11.1**F2,20 = 0.57NS
Number of adventitious stemsF1,20 = 0.54F2,20 = 17.5***NS
Figure 2.

Least squares means (±SD) for log transformed variables regressed on log(above-ground AFDW) (a) or log(total AFDW) (b–d) for control (white bars), sealed (shaded bars) and perforated (filled bars) treatments in Experiment I. Different letters indicate significant differences at the 5% level in t-tests between LSMs.

Experiment ii

Total DW was not affected by perforation (two-way anova; F1,65 = 2.80, P > 0.05, Table 3), although it was strongly affected by the main treatment (F2,65 = 56.6, P < 0.001, Table 3). No significant interaction was found between the main treatments and perforation (F2,65 = 0.92, P > 0.05). Although the water depth clearly affected both rhizome length and allocation to below-ground weight (t-tests between LSMs for shallow and deep water, P < 0.001), perforation did not affect these variables at any water depth (Tables 3 and 4) nor did it influence allocation to rhizome DW, root DW, aquatic root DW, number of stems or stem diameter (Table 4). The number of adventitious stems, however, increased clearly in perforated plants (Tables 3 and 4). The O2 concentration in the rhizomes was lower in the perforated treatment (two-way anova; F1,65 = 8.42, P < 0.01, Table 3), but neither the effect of the main treatments (F2,65 = 2.55, P > 0.05) nor the interaction term was significant (F2,65 = 2.25, P > 0.05).

Table 3.  Mean ± SD of variables measured in experiment II. Plants in each of the three main treatments (shallow water, deep water or in deep water with one stem per plant cut above the water surface) either had all stems perforated above the water surface (P) or the stems were left intact (C). Shoot weight is omitted from the ‘deep cut’ treatment, as one stem per plant was cut above the water surface and removed in this treatment
 ShallowDeepDeep Cut 
CPPCP
Total DW (g)15.6 ± 3.915.1 ± 4.29.4 ± 2.49.2 ± 1.68.8 ± 1.77.4 ± 2.0
Shoot DW (g)12.6 ± 3.012.3 ± 3.38.4 ± 2.38.2 ± 1.5  
Rhizome DW (g)1.12 ± 0.471.06 ± 0.300.36 ± 0.080.42 ± 0.070.59 ± 0.360.40 ± 0.14
Root DW (g)1.72 ± 0.561.51 ± 0.710.61 ± 0.180.55 ± 0.110.64 ± 0.240.53 ± 0.26
Aquatic root DW (g)0.15 ± 0.090.16 ± 0.110.03 ± 0.030.05 ± 0.020.06 ± 0.040.05 ± 0.04
Rhizome length (cm)172 ± 62163 ± 41 45 ± 1355 ± 1468 ± 27
Number of stems20.5 ± 4.520.5 ± 4.313.3 ± 2.413.9 ± 3.015.2 ± 4.213.8 ± 2.7
Number of adventitious stems1.3 ± 2.18.6 ± 6.50.8 ± 1.32.9 ± 2.71.2 ± 2.34.0 ± 3.5
Stem diameter (mm)2.3 ± 0.22.5 ± 0.22.1 ± 0.22.2 ± 0.22.2 ± 0.22.1 ± 0.2
O2 (%)18.9 ± 1.618.6 ± 1.818.1 ± 2.217.0 ± 2.119.1 ± 1.515.9 ± 3.6
Table 4.  Two-way ancovas for the effects perforation (P) and main treatments (M) in experiment II for different variables with log(above-ground DW) and log(total DW) as covariates (C). The dependent variables are log(x) transformed, except for aquatic roots and number of adventitious stems that are log(1 + x) transformed. Interaction terms involving the covariate are eliminated when non significant (NS, P > 0.05). *  P < 0.05, **  P < 0.01, ***  P < 0.001.
Dependent variableCovariate (C)TreatmentP × MC × PC × M
Perforation (P)Main (M)
  • a

    Term excluded as the interaction term (C × M) or (C × P) is significant.

Above-ground DW     
Below-ground DWF1,62 = 91.6**F1,62 = 1.37__aF2,62 = 0.70NSF2,62 = 5.61**
Total DW     
Rhizome DWF1,62 = 44.4***F1,62 = 0.02__aF2,62 = 1.72NSF2,62 = 4.67*
Root DWF1,62 = 146***F1,62 = 1.58__aF2,62 = 1.46NSF2,62 = 4.11*
Aquatic root DWF1,64 = 19.8***F1,64 = 1.01F2,64 = 4.59*F2,64 = 0.05NSNS
Rhizome lengthF1,64 = 20.4***F1,64 = 0.52F2,64 = 25.5***F2,64 = 1.37NSNS
Number of stemsF1,64 = 20.1***F1,64 = 0.09F2,64 = 4.09*F2,64 = 0.06NSNS
Stem diameterF1,64 = 2.03F1,64 = 2.19F2,64 = 2.52F2,64 = 1.64NSNS
Number of adventitious stemsF1,63 = 6.39*__aF2,63 = 1.61F2,63 = 0.78F1,63 = 4.51*NS

Experiment iii

The total DW was significantly lower for perforated plants (one-way anova; F1,22 = 11.60, P < 0.01, Table 5). Significant effects of the treatments were found on allocation to below-ground DW, rhizome DW, aquatic root DW, and on rhizome depth, rhizome length, number of rhizomes and number of adventitious stems (Table 6). No significant treatment effects were found for root DW, rhizome diameter, number of stems, stem diameter and shoot length (Table 6). There were no significant interactions between treatments and covariates for any variables. Comparisons between LSMs showed that perforated plants allocated less to below-ground DW, rhizome DW and to number of rhizomes but more to aquatic roots DW (Fig. 3a–d). LSMs also showed that new rhizomes were situated more superficially in the substrate and had shorter horizontal length (Fig. 3e, f). Number of adventitious stems was clearly higher in perforated plants although these plants were smaller (Table 5). The O2 concentration in the stem bases was significantly lower in perforated plants (one-way anova; F1,22 = 40.1, P < 0.001, Table 5). Flow rates up to 0.22 L h−1 stem−1 were measured in controls (mean ± S.D, 0.07 ± 0.06) while no flow was detected in perforated plants. At the beginning of the treatment, the redox potential in the substrate was 30 ± 28 mV. At harvest, the redox potential had increased to 105 ± 27 mV (pH varied between 7.2 and 8.6). There was no significant difference in redox potential between the treatments at the beginning (one-way anova; F1,23 = 0.005, P > 0.05) or at the end of the treatment (one-way anova; F1,23 = 0.09, P > 0.05).

Table 5.  Mean ± SD of variables in experiment III
 ControlPerforated
Total DW (g)70.3 ± 14.055.3 ± 8.3
Rhizome DW (g)10.7 ± 2.46.8 ± 1.7
Root DW (g)7.7 ± 1.36.4 ± 0.8
Aquatic Root DW (g)0.60 ± 0.300.76 ± 0.24
Number of stems39.1 ± 6.931.7 ± 5.1
Number of adventitious stems12.4 ± 6.022.9 ± 5.8
Number of new rhizomes10.4 ± 2.96.0 ± 2.1
Rhizome depth (cm)20.8 ± 1.917.9 ± 2.7
Rhizome length (cm)30.0 ± 5.219.1 ± 4.9
Rhizome diameter (mm)6.3 ± 0.65.6 ± 0.6
Stem diameter (mm)3.4 ± 0.33.1 ± 0.3
Stem length (cm)91.3 ± 9.590.0 ± 9.3
O2 (%)20.2 ± 1.314.4 ± 4.1
Table 6.  One-way ancovas for the effects of the treatments (perforated and control) in experiment III on different variables (dependent variable) with log(above-ground DW) or log(total DW) as covariates. The dependent variables are log(x) transformed, except for aquatic roots and number of adventitious stems that are log(1 + x) transformed. The interaction term between the dependent variable and the covariate was eliminated in all cases as it was non significant (P > 0.05). *  P < 0.05, **  P < 0.01, ***  P < 0.001
Dependent VariableCovariateTreatment
Above ground DW 
Below-ground DWF1,21 = 33.4***F1,21 = 14.9***
Total DW 
Root DWF1,21 = 16.4***F1,21 = 0.73
Rhizome DWF1,21 = 45.0***F1,21 = 10.9**
Aquatic root DWF1,21 = 15.0***F1,21 = 17.8***
Rhizome depthF1,21 = 0.0001F1,21 = 5.64*
Rhizome lengthF1,21 = 1.41F1,21 = 14.8***
Rhizome diameterF1,21 = 4.34F1,21 = 1.55
Number of rhizomesF1,21 = 6.09*F1,21 = 5.57*
Number of stemsF1,21 = 4.37*F1,21 = 1.75
Stem diameterF1,21 = 18.3***F1,21 = 0.04
Shoot lengthF1,21 = 0.73F1,21 = 0.06
Number of adventitious stemsF1,21 = 0.13F1,21 = 8.47**
Figure 3.

Least squares means (±SD) for log transformed variables regressed on log(above-ground AFDW) (a) or log(total AFDW) (b–f). Open bars represent control and filled bars represent the perforated treatment, in Experiment III.

Discussion

Evaluation of the method

Perforation of all stems above the water surface prevents pressurized ventilation in P. australis. There was no detectable flow in perforated stem cuttings and a considerable flow in intact stem cuttings. This is in accordance with results from Armstrong et al. (1996a) where culms with stem perforations caused by stem boring insects had considerably reduced flow rates compared to intact culms in situ. Decreased O2 concentrations were recorded in the rhizomes (experiment II) and the stem bases (experiment III) of perforated plants confirming the successful inhibition of pressurized ventilation. The flow rates in control plants were comparable to flow rates measured in situ by Armstrong et al. (1996a), when related to leaf sheath area (data not shown). Only the number of adventitious stems differed significantly between the sealed treatment in experiment I and the controls. Other effects of the perforation treatment can therefore be regarded as effects of reduced or absent pressurized ventilation. The greater number of adventitious stems in perforated treatments in all three experiments probably reflects a response to mechanical damage as previously shown in P. australis (van Deursen & Drost 1990). In experiment II, stem perforations were occasionally submerged as the water depth was increased. However, at the termination of the experiment, gas blown into the cut stems passed easily out of these holes, and no water was found in longitudinal sections of five internodes with submerged holes. The surface tension of the water and the small size of the holes probably prevented intrusion, and thus any blockage of the aeration pathway.

Effects on growth, biomass allocation and rhizome architecture

The decreased final total weight when pressurized ventilation was inhibited in experiment III clearly demonstrates that pressurized ventilation may improve the performance of P. australis. Pressurized ventilation also affected the allocation pattern between different plant parts. The decreased allocation to below-ground weight when pressurized ventilation was inhibited (experiments I and III but not experiment II) could be an adaptive response to low O2 concentrations in the below-ground parts, by increasing the ratio between O2-supplying and O2-consuming parts. Rhizomes function primarily as organs for vegetative dispersal and storage (Granéli et al. 1992), and since they are not essential for short-term survival, their production can be reduced. In contrast, roots are indispensable and allocation to root weight should therefore be maintained. This is in accordance with the results from experiments I and III, where the allocation to rhizomes but not to roots was affected by perforation. However, on a long-term basis, decreased allocation to rhizomes may severely affect the performance of emergent macrophytes in temperate regions, as energy stored in the rhizomes is needed for production of new shoots in spring (Granéli et al. 1992). This is especially important in deep water, where more structural tissue must be produced from resources derived from the rhizomes to enable that the shoots reach sufficient light to enable a net photosynthesis. Further, if the first cohort of stems is killed by, for example, frost or grazing, plants with decreased allocation to rhizomes might not be able to produce a second cohort. Thus, although decreased allocation to rhizomes may not be of immediate importance, it may have a strong effect on long-term survival in stochastic environments.

The restricted distribution of new rhizomes when pressurized ventilation was inhibited is in accordance with results on Typha domingensis Pers. (White & Ganf 1998) and indicates that O2 may, directly or indirectly, regulate rhizome architecture in emergent macrophytes. In new rhizomes lacking stems that have emerged above the water surface, the O2 concentrations in the growing apices are limited by the distance O2 must diffuse, i.e. from the junction between the old and the new rhizome to the new rhizome apex. Thus, when the O2 concentration at the junction between the old and new rhizome decreases due to inhibited pressurized ventilation, production of shorter rhizomes will be needed to maintain a critical O2 concentration in the rhizome apex. This is in accordance with in situ measurements of P. australis showing that the distribution of new rhizomes in the substrate is restricted in deep compared to shallow water (Weisner & Strand 1996) and that the O2 concentrations in stem bases decrease with increasing water depth (Yamasaki 1984; Weisner 1988). It has been suggested that anchoring difficulties may limit the distribution towards deep water of emergent macrophytes in soft sediments (Weisner 1991). As pressurized ventilation enables an increased allocation to below-ground parts and the production of long rhizomes situated deep in the sediment, thus improving the anchorage in the sediment, we suggest that species with pressurized ventilation may be able to grow in deeper water in areas with soft sediments than species which lack this ability. Plants with pressurized ventilation may also have an advantage over plants without this mechanism when new, unvegetated, areas are available for colonization, as longer and more numerous rhizomes could increase the rate of horizontal dispersal.

Factors affecting the importance of pressurised ventilation

No effects on plant growth of perforation were found in experiment II (plants growing in deep or shallow water) at either water depth, although the water depth had a clear impact on both growth and biomass allocation. This could possibly have been explained by a lack of throughflow ventilation as the plants were small and might not have developed shoots with different pressurizing capacity. However, a lower O2 concentration in perforated plants shows that perforation did affect O2 transport. Furthermore, no effect of perforation on plant growth was found in the treatment where one stem per plant was cut above the water surface to allow outflow. Thus, the lack of effects on plant parameters, except for adventitious stems, of the stem perforation in experiment II does not result from a lack of differences in rhizome O2 concentration but rather from a low O2 demand, caused by the low growth rate (in this case caused by aphid infestation) and the high redox potential of the substrate. The strong response to the perforation treatment in experiment III compared to experiment I can be explained by a combination of a low redox potential in the sediment, increasing the O2 demand of the below-ground tissues, and the fact that some stems were perforated in the controls, thus facilitating their pressurized ventilation. We therefore suggest that diffusion is sufficient to satisfy O2 demand in plants with a low growth rate or those growing in shallow water or in substrates with high redox potential; there may, however, be a shortfall in other situations.

Conclusions

Our experiments show that the ability to use pressurized ventilation may increase the performance of P. australis, albeit only under certain conditions. This is probably caused mainly by increased concentrations of O2 in roots and rhizomes, compared to those possible were O2 is supplied by diffusion alone. Pressurized ventilation probably allows P. australis to grow in deeper water and substrates with lower redox potential than would otherwise be possible.

Acknowledgements

We thank Jean Armstrong who kindly provided us with the T-shaped glass tube for flow rate measurements, Wilhelm Granéli, John Strand and anonymous referees for helpful comments on earlier versions of this paper and John Strand, Jonas Svensson and Helena Westerdahl for their assistance during some hectic days in the glasshouse. The study was supported by the European Community (4th Framework Programme – contract ENV4-CT95-0147 (EUREED II).

Received 24 June 1999 revision accepted 4 May 2000

Ancillary

Advertisement