• Waterlogging tolerance and its physiological basis were evaluated in Lophopyrum elongatum (a species from a salt marsh habitat), Triticum aestivum (cv. Chinese Spring), their amphiploid, and disomic chromosome addition lines.
• Growth in drained or waterlogged soil and aerated or deoxygenated stagnant nutrient solution were evaluated. Porosity, aerenchyma, rates of O2 consumption, radial O2 loss, and ethanolic fermentation in adventitious roots of selected genotypes were also measured.
•L. elongatum was more tolerant of deoxygenated stagnant nutrient solution or waterlogged soil (0–48% reductions in growth, respectively) than ‘Chinese Spring’ (63–84% reductions), the amphiploid (49–81% reductions), or the addition lines (56–92% reductions). Roots of L. elongatum had a higher constitutive porosity than ‘Chinese Spring’, resulting in greater internal O2 movement in these roots when intact plants were first transferred to an O2-free root medium. This trait, as well as the slower growth and development of L. elongatum, might have contributed to its greater waterlogging tolerance.
• Despite the differences between L. elongatum and ‘Chinese Spring’ in some traits, none of the Lophopyrum × wheat lines showed substantial improvement in waterlogging tolerance.
Waterlogging can cause severe yield reductions in wheat crops throughout the world (Cannell et al., 1980; van Ginkel et al., 1992; Musgrave, 1994). The intolerance of wheat to waterlogging presumably results from its lacking, or having only a low expression of, traits associated with tolerance to waterlogging (Huang et al., 1997; Thomson et al., 1992; Hoffman et al., 1993). Undomesticated species related to wheat, known as ‘wild relatives’, may possess traits associated with tolerance to waterlogging, and some species may be suitable as donors of these traits into wheat. Introgression of the genome of Lophopyrum elongatum, a species that inhabits salt marshes, may be one source of germplasm for improving waterlogging tolerance in wheat (Taeb et al., 1993).
Waterlogged soils are usually anaerobic and can also be chemically reduced (Laanbroek, 1990). Many tolerant species develop extensive adventitious root systems in response to soil waterlogging (Jackson & Drew, 1984; Laan et al., 1989). Root penetration into anaerobic substrates is dependent upon O2 movement from the atmosphere to the root apex, via shoot and root aerenchyma (Armstrong, 1979). Hence, wetland species with more aerenchyma and a lower resistance to internal O2 diffusion produce longer adventitious roots in waterlogged soil than intolerant species with less aerenchyma (Armstrong, 1979; Justin & Armstrong, 1987). The longitudinal diffusion of O2 in a root is also affected by its rate of loss in the radial direction from the root to the soil (Armstrong, 1979). Some wetland species have reduced rates of radial O2 loss (ROL) from the basal regions of their adventitious roots, thus enhancing longitudinal O2 movement towards the root apex (Armstrong, 1964; Armstrong, 1979).
Internal aeration greatly improves the O2 status in roots of waterlogged plants but some tissues, such as root apices (densely packed, with higher metabolic rates and most distant from the O2 source), may still suffer O2 deficiencies (Waters et al., 1991a). In the absence of O2, oxidative phosphorylation is inhibited and fermentation pathways are necessary to maintain glycolysis and therefore some production of ATP, although only at a fraction of the rate possible by oxidative phosphorylation (Vartapetian & Jackson, 1997). The most common end product of fermentation in wetland plants is ethanol (Davies, 1980; Smith et al., 1986) and fast rates of ethanolic fermentation would result in higher ATP production, as long as a supply of substrate remained available.
The poorly developed adventitious root system and relatively low root porosity of hexaploid wheat are thought to contribute to its sensitivity to waterlogging (Thomson et al., 1992). However, addition of the L. elongatum genome and, in some cases, individual chromosomes into hexaploid wheat increased the penetration of adventitious roots into waterlogged soil (Taeb et al., 1993). This earlier result suggests that L. elongatum has traits that increase the longitudinal diffusion of O2 within its adventitious roots and that these features are heritable in crosses with hexaploid wheat. The experiments described in the present study evaluated growth, porosity, aerenchyma, rates of O2 consumption, ROL and ethanolic fermentation in adventitious roots of Triticum aestivum cv. ‘Chinese Spring’, L. elongatum, and a L. elongatum × Chinese Spring amphiploid. The growth and root porosity of the set of disomic chromosome addition lines of L. elongatum into ‘Chinese Spring’ were also assessed for plants grown in aerated or deoxygenated stagnant nutrient solution as was growth of the lines in drained or waterlogged soil.
Materials and Methods
Triticum aestivum L. cv. ‘Chinese Spring’, 2n = 6x = 42, genome AABBDD; Lophopyrum elongatum Löve, 2n = 2x = 14, genome EE; the amphiploid of ‘Chinese Spring’ × Lophopyrum elongatum, 2n = 8x = 56, genome AABBDDEE; and the disomic addition lines with each of the seven L. elongatum chromosome pairs individually incorporated into ‘Chinese Spring’ (designated DA1E; disomic addition of 1E into ‘Chinese Spring’ and so on, as described by: Dvořák & Knott, 1974; Dvořák, 1980; Hart & Tuleen, 1983; Dvořák & Chen, 1984) were supplied by J. Dvořák (University of California, Davis, CA, USA). The accession of L. elongatum used in this study was designated D (originally from the University of Manitoba, Winnipeg, Canada); this is not the same as the accession (8A206) used in the production of the amphiploid (Rommel & Jenkins, 1959) as the later accession was not available.
Confirmation of chromosomic constitution of the lines using genomic in situ hybridization
Genomic DNA from Lophopyrum elongatum (EeEe; Accession H6695) and hexaploid wheat (cv. Tjalve) was used as probes. Probe labelling was carried out using nick translation and the fluorescent nucleotides rhodamine-4-dUTP or fluorescein-11-dUTP (Amersham) following the protocol by Anamthawat-Jónsson et al. (1996).
Chromosome squash preparations of root tips were made according to Schwarzacher & Leitch (1994) using 24 h ice-water treatment to arrest chromosomes in mitosis followed by fixation in 3 : 1 (v/v) ethanol : glacial acetic acid. Chromosome counts were taken for at least three plants of each addition line, and genomic in situ hybridization was evaluated in at least one of these three plants. The in situ hybridization followed Anamthawat-Jónsson & Reader (1995) and Anamthawat-Jónsson et al. (1996). The experiments were carried out with combined denaturation of both chromosomes and probe at 87°C for 10 min using a modified thermocycler with a simulated slide control function (Hybaid) (Anamthawat-Jónsson et al., 1996). 4,6-Diamino-2-phenylindole (DAPI) was used as the counter-stain. Fluorescence was examined using an epifluorescence microscope with appropriate filters.
Experiments on ‘Chinese Spring’, L. elongatum and a Chinese Spring × L. elongatum amphiploid
A preliminary experiment showed that the various genotypes had different developmental rates – that of L. elongatum being much slower than that of ‘Chinese Spring’, and the amphiploid intermediate. Therefore, germination of the lines was staggered to ensure plants were at the same developmental stage (2.0–2.2 leaves) on the date coinciding with the onset of treatments (ages are given below). Seeds were washed in 0.4% NaHClO− solution for 45 s and rinsed thoroughly with deionized water before being placed on plastic mesh floating on 10% strength aerated nutrient solution in darkness in a 20 : 15°C day : night phytotron. The composition of the nutrient solution at full strength was: (mol m−3) K+ 3.95, Ca2+ 1.50, Mg2+ 0.40, NH4+ 0.625, NO3− 4.375, SO42− 1.90, H2PO4− 0.20, Na+ 0.20, H4SiO4− 0.10; and the micronutrients (mmol m−3) Cl 50, B 25, Mn 2, Zn 2, Ni 1, Cu 0.5, Mo 0.5, Fe-EDTA 50. The solution also contained 2.5 mol m−3 MES and the pH was adjusted to 6.5 with KOH (increasing the final K+ concentration to 5.60 mol m−3). All chemicals used were analytical grade.
Three d after imbibition (7 d in the case of L. elongatum) the seedlings were exposed to natural light and the concentration of the nutrient solution was increased to 25% strength. Five days after imbibition (14 d in the case of L. elongatum) the seedlings were transplanted into 4.5 l plastic pots containing aerated full strength nutrient solution. Four seedlings of the same genotype were transplanted into each pot and held into the lid using polystyrene foam holders that allowed expansion as the plants grew. A total of six pots of each genotype were established. Both the lid and the pot were covered with aluminium foil to prevent light penetration into the pot. Root-zone O2 treatments were initiated at the 2.0–2.2 leaf stage as the first whorl of adventitious roots started to emerge (plant ages were: ‘Chinese Spring’, 12 d; Amphiploid, 15 d; L. elongatum, 29 d). Three of the pots containing each genotype continued to be well aerated while the other three pots were given a hypoxic pretreatment by flushing with N2 gas until the O2 concentration in solution reached approx. 0.03 mol m−3 after which the pots were left stagnant for 24 h. Following this pretreatment the solution in these pots was replaced with deoxygenated nutrient solution containing 0.1% w/v agar to prevent convective movements in the solutions in those pots (‘stagnant’ treatment) (Wiengweera et al., 1997). Solutions in aerated pots were also renewed at this time and subsequently all solutions were renewed every 7 d, and in the middle of each 7-d period the stagnant pots were flushed with N2 gas for 10–20 s to ensure the agar remained suspended and to disrupt the development of boundary layers within the rhizosphere. O2 concentrations in the bulk medium in stagnant pots never rose above 0.003 mol m−3 (Syland O2 meter, Model 610). O2 concentrations in aerated pots at 20°C averaged 0.28 mol m−3. Treatments were maintained for 21 d and pots were arranged in a randomised complete block design with three replicates.
An initial harvest was taken at the time the treatments were imposed. Shoot and root d. wt were measured and plant developmental stages recorded. Following the 21 d of treatments the lengths of the longest adventitious roots, and d. wt of shoot, seminal and adventitious roots were measured on each plant. Relative growth rates (RGR) were calculated from the initial and final harvests on a d. wt basis. In an additional experiment maximum adventitious root lengths were measured every 7 d until there were no further increases in root length for 14 d. Three replicates of each genotype × aeration treatment were used.
Rates of O2 consumption by adventitious roots
Rates of O2 consumption by adventitious roots were measured polarographically at 20°C using the technique and apparatus similar to that described by Lambers & Steingröver (1978). Adventitious roots (70–120 mm in length) were excised from plants grown in aerated or stagnant nutrient solution. The 10 mm apical region of each root was excised and the expanded root tissues were used for the measurements that commenced in an air-saturated nutrient solution (same composition as culture solution given above). The segments of adventitious roots used did not have laterals.
Hand cross sections were taken at 10 mm below the root-shoot junction and 50 mm behind the apex of adventitious roots of plants grown in stagnant deoxygenated nutrient solution that had reached the maximum lengths. The sections were photographed using a photographic-microscope. The aerenchyma in each cross section was measured by tracing around the area of aerenchyma in the photograph, cutting out that area and weighing it and expressing that as a percentage of the weight of the area of the total cross section.
Radial O2 loss (ROL) measurements
Plants were grown as described above and exposed to treatments for 21–25 d before being transferred to a 20°C controlled temperature room. The shoot base of each plant was held in a rubber lid, and the lid was fitted on a 100 mm × 100 mm × 170 mm (w × b × h) Perspex chamber filled with an O2-free solution containing 0.1% (w/v) agar and in mol m−3; Ca2+ and SO42− at 0.5, and K+ and Cl− at 5. The root system was immersed in the O2-free solution while the shoot system was exposed to air. A cylindrical platinum electrode (ht. 5 mm; id 2.25 mm) was placed around a selected adventitious root with a length of 100–120 mm. Rates of O2 loss from the root to the electrode were measured (Armstrong & Wright, 1975; Armstrong, 1994) with the centre of the electrode positioned at 50, 40, 30, 20, 10 and 5 mm behind the root apex. One root was measured per plant for four replicate plants of each genotype and treatment.
In vivo ethanol production rates in adventitious root tips
Pots of plants raised in aerated nutrient solution for 28 d (‘Chinese Spring’ and the amphiploid) or 42 d (L. elongatum) were transferred to a 20°C controlled temperature room. The seminal root system of each plant was excised and the plants with intact adventitious roots were placed in nutrient solution (composition given above with addition of 20 mol m−3 glucose and 10 mg l−1 carbenicillin). This solution composition was used throughout the remainder of this experiment. The roots were ‘hypoxically pretreated’ for 18 h by flushing the solutions with an air : N2 gas mixture so that the dissolved O2 concentration in the solution was 0.020 mol m−3. The intact shoots were enclosed in a plastic bag and flushed with N2 gas to prevent internal O2 movement during the hypoxic pretreatment. After the pretreatment, adventitious root tips (10 mm) were excised and placed into nutrient solution at 0.020 mol m−3 O2 for 5 h to ‘rest’ the tissues. The tips were then transferred into 30 ml crimp top vials containing 5 ml of O2-free nutrient solution and the vials were sealed and flushed with high purity N2 gas for 20 min to ensure anaerobiosis, after which the outlet and inlet needles were removed from the self-sealing cap. The vials were submerged in O2-free water at 20°C contained in a larger vial and placed on an orbital shaker so that the tips were incubated whilst gently shaking. After 2 h, each vial was opened and the incubation solution was transferred to a new vial, capped and kept on ice. Ethanol in the tips was extracted with ice-cold 5% (w/v) perchloric acid and the extract was neutralized with K2CO3 before being assayed for ethanol. Ethanol in the incubation solution and the neutralized extracts was assayed using the enzymatic technique of Beutler (1983). Recovery of ethanol from the incubation solution was 96% and from the tissue extracts it was 67%. Ethanol in blank solutions from vials incubated without root tips was below the detection limits of 1.6 µmol l−1 for the technique.
The data on the growth and development and rate of adventitious root O2 consumption were subjected to an ANOVA to examine the effects of genotype, aeration treatment and their interaction. The data on aerenchyma formation were subjected to a split-plot ANOVA (with root as main plot and section position as subplot) to examine the effects of genotype and section position on aerenchyma proportions. The data on ROL were subjected to a split-plot ANOVA (with root as main plot and electrode position as subplot) to examine the effects of genotype, treatment, electrode position and their interactions. Data on adventitious root lengths within each treatment and rate of ethanol production were subjected to an ANOVA to examine the effect of genotype.
Experiments on T. aestivum cv. ‘Chinese Spring’, L. elongatum, a ‘Chinese Spring’ × L. elongatum amphiploid and the seven disomic chromosome addition lines
L. elongatum and the amphiploid were more tolerant to growth in stagnant solution than ‘Chinese Spring’ (see Results). Further experimentation was therefore conducted on the growth and root porosity of the L. elongatum×‘Chinese Spring’ chromosome addition lines in stagnant solution and waterlogged soil.
Plants were grown in solution culture as described above. Root-zone O2 treatments were initiated at the 2.0–2.2 leaf stage as the first whorl of adventitious roots started to emerge (plant ages were: ‘Chinese Spring’, DA1E, DA2E, DA3E, DA4E, DA6E, 12 d; DA7E, 13 d; DA5E, 14 d; Amphiploid, 15 d; L. elongatum, 29 d). Treatments were maintained for 21 d and pots were arranged in a randomised complete block design with three replicates.
Seeds were washed in 0.4% NaHClO− solution for 45 s and rinsed thoroughly with deionized water before being germinated in darkness on plastic mesh floating on aerated 0.5 mM CaSO4 solution for 24 h (72 h in the case of L. elongatum). The germination and subsequent planting into soil was staggered so that the plants were at the same developmental stage (2.0–2.2 leaves) at the onset of the waterlogging treatment. The seeds were then planted into soil in pots (450 mm high and 150 mm in diameter) that contained c. 7 kg of a fine grey sand from the top 150 mm of a lateritic podzolic soil from near Esperance, Western Australia. Nutrients mixed through the soil before planting were: (mg kg−1) NH4NO3, 357; KH2PO4, 60; K2SO4, 160; CaCl2·2H2O, 178; MgSO4·7H2O, 22; ZnSO4·7H2O, 11; MnSO4, 11; CuSO4·5H2O, 6; H3BO3, 0.8; Na2MoO4·2H2O, 0.2. Four seeds per pot were planted and these were later thinned to two seedlings per pot. Pots were watered daily to field capacity. At the 2.0–2.2 leaf stage (plant ages were: ‘Chinese Spring’, DA1E, DA2E, DA3E, DA4E, DA6E, 11 d; DA7E, 12 d; DA5E, 13 d; Amphiploid, 14 d; L. elongatum, 27 d) waterlogging was imposed in half of the pots by partially submerging them in tanks of deionized water so that it entered the bottom of the pot and rose to the soil surface. A transparent plastic tube attached to the drainage hole and vertically along the side of the pot prevented water drainage and allowed monitoring of the water level within the soil column. The waterlogging ranged from level with the soil surface to 3 mm above the surface. Controls, in freely draining pots, were watered to field capacity each day. The treatments were maintained for 21 d. Pots were arranged in a randomised complete block design with three replicates in a 20 : 15°C (day : night) phytotron.
An initial harvest was taken at the time the treatments were imposed in both experiments. Shoot and root d. wt were measured and plant developmental stages recorded. Following the 21 d of treatments the numbers of tillers and adventitious roots, length of the longest adventitious root, and shoot, seminal and adventitious root d. wt were measured on each plant. RGR were calculated from the initial and final harvests on a d. wt basis.
Porosity (% gas volume/root volume) was measured for adventitious and seminal roots of plants grown in aerated or stagnant nutrient solutions using the method of Raskin (1983) with equations as modified by Thomson et al. (1990). The root systems of two plants from each pot were separated into adventitious and seminal roots and cut into 40–50 mm segments, and a subsample of c. 2 g f. wt was used for the measurements.
The values of shoot d. wt, shoot RGR and adventitious and seminal root d. wt under aerated or drained conditions were subjected to an ANOVA to detect the effect of genotype. The values of these variables under stagnant or waterlogged conditions were expressed as a percentage reduction in comparison with the values under aerated or drained conditions, respectively, and these data were subjected to an ANOVA to detect the effect of genotype. The remaining data on aspects of root and shoot development were subjected to an ANOVA to determine the main effects of genotype and treatment and the interaction between the two. Root porosity data were subjected to an ANOVA to determine the effects of genotype, treatment and root type and their interactions. Means were compared using LSDs (P = 0.05).
Genomic in situ hybridization confirmed the chromosome constitution of all addition lines, except DA1E. The probe from L. elongatum hybridized to one pair of chromosomes in all addition lines. All addition lines (except some DA1E plants) contained 44 chromosomes. To further characterize the status of DA1E, chromosome counting was performed on preparations made from an additional six plants. Three of these had 44 chromosomes while the others had only 41, 42 or 42 plus telomere. One of the DA1E plants with 44 chromosomes and one with 42 chromosomes were examined using in situ hybridization; both had one pair of Lophopyrum chromosomes, so one pair of wheat chromosomes must have been missing in the plant with 42 chromosomes. Data for DA1E are therefore given in brackets in the tables so as to remind the reader that, unlike all the other lines used in this study, the means given for DA1E may contain measurements on some plants that did not contain the full set of chromosomes.
Shoot d. wt differed among the genotypes when grown in aerated solution (Table 1); the values for ‘Chinese Spring’ and the amphiploid were 7.2- and 5.2-times greater than those for L. elongatum, respectively. When ‘Chinese Spring’ and the amphiploid were grown in stagnant solution, shoot d. wt were reduced by 63% and 49% in comparison with aerated controls, respectively; but shoot wt was not reduced in L. elongatum. The shoot RGR of ‘Chinese Spring’ was 1.2-fold higher than that of L. elongatum when grown in aerated solution, and the RGR of the amphiploid was between its parents (Table 1). When ‘Chinese Spring’ and the amphiploid were grown in stagnant solution, shoot RGR was reduced by 28% and 21% in comparison with aerated controls, respectively; but shoot RGR was not reduced in L. elongatum.
Table 1. Aspects of shoot and root growth of Triticum aestivum cv. Chinese Spring, Lophopyrum elongatum and their amphiploid after being grown until the 2.0–2.2 leaf stage in aerated solution and then for 21 d in aerated or stagnant deoxygenated solution. Values given are means of 3 replicates. LSDs are for the genotype × treatment interaction
Shoot d. wt (g)
Shoot RGR (mg g−1 d−1)
Root d. wt (g)
Level of significance: ns, not significant; **P < 0.01; ***P < 0.001.
In aerated solution, the seminal root d. wt of ‘Chinese Spring’ and the amphiploid was 4.3- and 3.7-fold higher than that of L. elongatum (Table 1). The seminal root d. wt of ‘Chinese Spring’, the amphiploid and L. elongatum grown in stagnant solution was reduced by 81%, 76% and 55%, respectively, in comparison with aerated controls. L. elongatum produced less adventitious root d. wt than ‘Chinese Spring’ and the amphiploid in aerated solution, but the genotype × treatment interaction was not significant for this character (Table 1).
Aerenchyma and O2 consumption rates as related to lengths of adventitious roots grown in stagnant deoxygenated solution
The maximum length of adventitious roots of the amphiploid in stagnant solution was 40 mm greater than those of ‘Chinese Spring’ or L. elongatum. The latter two genotypes did not differ (Table 2). Aerenchyma in these longest roots constituted between 34 and 42% of the cross sectional root area of all three genotypes when grown in stagnant solution and there was no effect of genotype (P = 0.360). There was a significant genotype × section position effect, with L. elongatum having more aerenchyma at 10 mm below the root/shoot junction than ‘Chinese Spring’, but less at 50 mm behind the root tip (Table 2). Rates of O2 consumption by expanded tissues of adventitious roots of the three genotypes showed no significant genotype (P = 0.418), treatment (P = 0.095) or genotype × treatment effects (P = 0.250). Thus, the values given in Table 2 are the means for roots of each genotype grown in aerated and stagnant solutions. Furthermore, it made no difference to comparisons if the results were expressed on a volume or a f. wt basis, or whether roots were vacuum infiltrated with nutrient solution before the measurements (data not shown).
Table 2. The cross sectional area of aerenchyma at 10 mm below the root/shoot junction and 50 mm behind the tip of the longest adventitious roots of Triticum aestivum cv. ‘Chinese Spring’, Lophopyrum elongatum and their amphiploid in stagnant solution. The LSD refers to the genotype × section position effect. The maximum adventitious root length is that beyond which no further elongation was observed for 14 d in stagnant deoxygenated solution. The predicted maximum root length possible in an O2-free medium was calculated using the model of Armstrong (1979) as described in the text. Values are the means of three replicates. The LSDs refers to genotype × root position effect (aerenchyma percentage) and genotype effect (rate of O2 consumption and measured root length data)
Rate of O2 consumption (nmol O2 g−1 f. wt s−1)
Maximum length of adventitious roots (mm)
10 mm below root/shoot
50 mm behind tip
Level of significance: ns, not significant; *P < 0.05; ***P < 0.001.
In order to examine further the factors determining maximum lengths of roots of plants grown in the deoxygenated stagnant solution and solely dependant on internal O2 diffusion to supply the apex, the model of Armstrong (1979) was used to predict the theoretical maximum adventitious root lengths (ι) for the three genotypes when grown under these conditions. The model is described by the equation:
(Do, O2 diffusivity in air (2.01 × 10−1 cm2 s−1 at 20°C); τ, tortuosity factor, which is assumed to be 1; Co, O2 concentration in moist air (2.728 × 10−4 g cm−3 at 20°C); ɛ, the fractional root porosity (data from Table 2); M, rate of O2 uptake by the adventitious root tissue (data from Table 2) expressed as g O2 cm−3 s−1).
The theoretical maximum root lengths of the three genotypes in an O2-free root medium, calculated using the model, were similar to the measured values (Table 2). Although genotypic differences in aerenchyma and root O2 consumption rate were not significant as main effects (Table 2), the slightly greater aerenchyma and lower rate of O2 consumption by root tissues in the amphiploid interacted to produce the longer maximum root length in the amphiploid. When the individual contributions of the larger amount of aerenchyma and lower rate of O2 consumption on root length were analysed using the model, each accounted for an 11 and 26 mm increase, respectively, relative to the theoretical maximum root length value for ‘Chinese Spring’.
ROL profiles along adventitious roots, when in an O2-free medium
The steep declines in ROL from the base to the apex of adventitious roots of plants raised in aerated solution before being transferred to the anoxic medium contrast with the shallow decline, if any, in the ROL profiles along roots of plants raised in stagnant solution (Fig. 1). This difference resulted in a significant treatment × distance behind apex interaction (P < 0.001). For plants grown in aerated solution before the measurements, adventitious roots of L. elongatum had a higher rate of ROL than those of ‘Chinese Spring’ or the amphiploid (P < 0.001). The mean rates of ROL (i.e. the mean of all positions measured along the roots) from the adventitious roots of ‘Chinese Spring’ and the amphiploid grown in stagnant solution were faster than those from plants raised in aerated solution (genotype × treatment, P < 0.001). By contrast, the mean rates of ROL from the adventitious roots of L. elongatum raised in aerated solution did not differ from those of plants grown in stagnant solution, although rates near the tip had increased while those in the most basal positions had decreased. There was no genotype × treatment × distance behind apex effect (P = 0.982) amongst the genotypes.
In vivo ethanol production rates in anoxic adventitious root tips
The rates of ethanol production in the 10 mm tips of hypoxically pretreated adventitious roots were not different among the three genotypes. The rates were (nmol g−1 f. wt s−1): ‘Chinese Spring’ 0.92 ± 0.08, L. elongatum 0.79 ± 0.11, and amphiploid 1.14 ± 0.07 (P = 0.08).
Experiments on waterlogging tolerance in T. aestivum cv. ‘Chinese Spring’, L. elongatum, a ‘Chinese Spring’ ×L. elongatum amphiploid and disomic addition lines
Solution culture experiment
Growth and development
The values of shoot d. wt, shoot RGR, adventitious and seminal root d. wt of L. elongatum were less than that of all the other genotypes in aerated solution and L. elongatum was proportionately less affected by growth in stagnant solution than ‘Chinese Spring’ (Table 3). The shoot d. wt of L. elongatum grown in stagnant solution was 39% less than that in aerated solution. The shoot d. wt of ‘Chinese Spring’, the amphiploid and chromosome addition lines were reduced by between 54 and 68% by growth in stagnant solution with none of the genotypes being less reduced than ‘Chinese Spring’. The shoot RGR of all the genotypes was reduced by around 20–30% by growth in stagnant solution. The shoot RGR of DA4E was 7% less reduced by growth in stagnant solution than that of ‘Chinese Spring’. The adventitious root d. wt of the amphiploid was 23% less reduced by growth in stagnant solution than that of ‘Chinese Spring’. Growth in stagnant solution reduced seminal root d. wt to between 95 and 97% of that in aerated solution for ‘Chinese Spring’, the amphiploid and the chromosome addition lines. In contrast, the seminal root d. wt of L. elongatum was reduced by only 19%.
Table 3. Shoot d. wt, shoot RGR, adventitious and seminal root d. wt of Triticum aestivum cv. ‘Chinese Spring’, and the chromosome addition lines and amphiploid of Lophopyrum elongatum into ‘Chinese Spring’ after being grown in aerated nutrient solution until the 2.0–2.2 leaf stage and then either a further 21 d in aerated solution or stagnant deoxygenated solution. Values are the means of three replicates for aerated plants and percentage reductions for plants in the stagnant treatment. LSDs are for the effect of genotype
In aerated solution the number of tillers produced by L. elongatum was c. 50–70% fewer than the wheat lines (Table 4). There was a decrease of between 25 and 45% in the number of tillers for all the genotypes when grown in stagnant solution. Although proportionately reduced to a similar extent as in ‘Chinese Spring’, the absolute decrease in the number of tillers of L. elongatum was not significant (Table 4).
Table 4. The number of tillers and the number of adventitious roots per stem (tillers + mainstem) of Lophopyrum elongatum, Triticum aestivum cv. ‘Chinese Spring’, their amphiploid, and chromosome addition lines grown in aerated solution until the 2.0–2.2 leaf stage and then in aerated or deoxygenated stagnant nutrient solution for 21 d. Values are the means of three replicates. LSD’s are for the genotype × treatment interaction
The number of adventitious roots produced by the plants is considered on a per stem (main stem + tillers) basis to avoid the association between adventitious root and tiller formation (Klepper et al., 1984). In aerated conditions the amphiploid produced fewer adventitious roots per stem than ‘Chinese Spring’ and L. elongatum, between which there was no difference (Table 4). None of the addition lines produced more adventitious roots per stem than ‘Chinese Spring’ in either aerated or stagnant solution. Growth in stagnant solution increased the number of adventitious roots per stem (mean values: aerated 2.38, stagnant 3.69, LSD 0.16)
In aerated solution the adventitious root porosity in both L. elongatum and the amphiploid was greater than that in ‘Chinese Spring’, but none of the addition lines differed from ‘Chinese Spring’ (Table 5). Plants grown in stagnant solution had a greater adventitious root porosity than those grown in aerated solution (mean values: aerated 4.86, stagnant 18.35, LSD 0.64) but the proportional increase differed among genotypes. The adventitious root porosity of L. elongatum increased by around 30% when grown in stagnant conditions, whereas increases in the other genotypes ranged from 200 to 700%. The amphiploid had a higher adventitious root porosity under stagnant conditions than either ‘Chinese Spring’ or L. elongatum. When grown in stagnant solution, (DA1E) also had a higher adventitious root porosity than ‘Chinese Spring’, but not L. elongatum; none of the other addition lines differed from either ‘Chinese Spring’ or L. elongatum. The seminal root porosity of L. elongatum was higher than that of any other genotype under aerated or stagnant conditions. With the exceptions of L. elongatum and (DA1E), growth in stagnant solution did not increase the seminal root porosity.
Table 5. Porosity of seminal and adventitious roots of Lophopyrumelongatum, Triticum aestivum cv. ‘Chinese Spring’, their amphiploid, and chromosome addition lines grown in aerated solution until the 2.0–2.2 leaf stage and then in aerated or stagnant deoxygenated solution for 21 d. Values are the means of three replicates. The LSD is for the genotype × root type × treatment interaction
The values of shoot d. wt, shoot RGR, adventitious and seminal root d. wt of L. elongatum were less than those of all the other genotypes when grown in drained soil (Table 6). With the exception of shoot RGR, L. elongatum was proportionately less affected by growth in waterlogged soil than all other genotypes. None of the L. elongatum addition lines were less affected by waterlogging than ‘Chinese Spring’; reductions being 81–92% in shoot d. wt, 36–51% in shoot RGR, 71–87% in adventitious root d. wt, and 96–98% in seminal root d. wt in comparison with the respective drained controls.
Table 6. Shoot d. wt, shoot RGR, adventitious and seminal root d. wt of Lophopyrum elongatum, Triticum aestivum cv. ‘Chinese Spring’, their amphiploid, and chromosome addition lines after being grown in drained soil until the 2.0–2.2 leaf stage and then either a further 21 d in drained or waterlogged soil. Values are the means of three replicates for plants in drained soil and percentage reductions for plants in waterlogged soil. LSDs are for the effect of genotype
Shoot d. wt (g)
Shoot RGR (mg g−1 d−1)
Root d. wt (g)
Level of significance: ns not significant; *P < 0.05; **P < 0.01; ***P < 0.001.
In drained conditions the lengths of the longest adventitious roots of all genotypes were greater than those of L. elongatum (Table 7). Apart from L. elongatum, only DA2E differed significantly from ‘Chinese Spring’, producing shorter roots. Waterlogging reduced the mean length of the longest adventitious root from 356 mm to 125 mm (LSD 19 mm). The lengths of the longest adventitious roots of L. elongatum were proportionately less reduced than those of the other lines. ‘Chinese Spring’, the amphiploid and DA3E had longer roots in waterlogged soil than those of L. elongatum. None of the genotypes produced roots longer than those of ‘Chinese Spring’, although roots of DA5E were significantly shorter.
Table 7. Length of the longest adventitious roots and number of adventitious roots per stem (tillers plus main stem) of Lophopyrum elongatum, Triticum aestivum cv. ‘Chinese Spring’, their amphiploid, and chromosome addition lines grown until the 2.0–2.2 leaf stage in drained soil and then for a further 21 d in pots of drained or waterlogged soil. Values are the means of three replicates. LSDs are for the effect of genotype (root length data) and genotype × treatment interaction (root number data)
The genotypes did not differ in the number of adventitious roots produced per stem under drained conditions (Table 7). Waterlogging increased the number of adventitious roots per stem, from 2.53 to 7.25 (LSD 0.34). This increase was not uniform amongst genotypes: in L. elongatum the number of adventitious roots per stem did not differ between the drained and the waterlogged treatment (Table 7). All genotypes, with the exception of the amphiploid, produced more adventitious roots per stem than L. elongatum in waterlogged conditions, but only DA4E produced more than ‘Chinese Spring’.
L. elongatum tolerated growth in stagnant solution or waterlogged soil better than ‘Chinese Spring’. Reductions in shoot d. wt of L. elongatum were between 13 and 39% in stagnant solution and 48% in waterlogged soil (Tables 1, 3 and 6). In an earlier study shoot d. wt of L. elongatum was reduced by c. 45% in waterlogged compost (Taeb et al., 1993). The shoot d. wt of ‘Chinese Spring’ was reduced by 50–60% in stagnant solution and by c. 80% in waterlogged soil (Tables 1, 3 and 6). Several physiological parameters were assessed in the present study, and the seminal root porosity and constitutive porosity in adventitious roots was higher in L. elongatum than ‘Chinese Spring’. ROL from adventitious roots of L. elongatum grown in aerated solution before being transferred to an O2-free root medium was also substantially higher, than for ‘Chinese Spring’, presumably due to the high constitutive porosity in the roots of L. elongatum. However, L. elongatum also differed greatly in growth habit compared with ‘Chinese Spring’; its shoot RGR and development were much slower. A low shoot RGR may offer an ecological advantage to plants found in stressful environments (Grime & Hunt, 1975; Lambers & Poorter, 1992). The differences in habit between L. elongatum and T. aestivum (cv. ‘Chinese Spring’) complicate comparisons of stress tolerance between these two species and the contribution made by this habit to the apparent waterlogging tolerance in L. elongatum is unknown.
The amphiploid was also more tolerant to growth in stagnant solution than ‘Chinese Spring’ (Table 1) although this difference was not always significant (Table 3). The slightly enhanced tolerance in the amphiploid may result from it having an inherently slower growth rate than ‘Chinese Spring’ (Tables 1 and 3) and/or differences in aspects of adventitious root physiology (Table 2). However, there was no difference in tolerance to growth in waterlogged soil between ‘Chinese Spring’ and the amphiploid (Table 6).
The values of porosity (17–22%) and cross-sectional areas of aerenchyma (35–40%) in adventitious roots of the genotypes grown in stagnant solution in the present study were equal to, or greater than, values reported for other wheat genotypes. The porosity of the adventitious roots of waterlogging sensitive genotypes, such as Coker 9835 (Huang & Johnson, 1995) and Seri-82 (Boru, 1996) in hypoxic solution was c. 10%, and for less sensitive genotypes such as Jackson (Huang & Johnson, 1995) and Parula/Sara (Boru, 1996) porosity was c. 20%. Aerenchyma occupied 19% in cross sections from the adventitious roots of the waterlogging sensitive genotype Bayles and 30% in the less sensitive Savannah, when grown in waterlogged sand culture (Huang et al., 1994). As such ‘Chinese Spring’ is equal, or better than, other wheat genotypes in its capacity to form aerenchyma.
The higher constitutive porosity in adventitious roots of L. elongatum raised in aerated solution enabled O2 movement to the apex of 100–120 mm roots upon transfer of intact plants to an O2-free root medium. By contrast, the low porosity in adventitious roots of ‘Chinese Spring’ and the amphiploid raised in aerated solution was insufficient to support internal O2 diffusion to the apex of roots of a similar length. Aerenchyma formation induced by waterlogging, or ethylene treatments, occurs over a period of days (Konings, 1982; He et al., 1996), so that constitutive aerenchyma may allow for internal O2 diffusion in the roots during this time.
The porosity in the adventitious roots of all three genotypes was higher for plants raised in stagnant solution compared with those in aerated solution (Table 5), and this enhanced the internal movement of O2 to the apical regions of roots when in an O2-free medium (Fig. 1). Nevertheless, unlike many monocotyledonous wetland species, ROL from the basal regions of roots from the three genotypes remained high, indicating the absence of a strong barrier to ROL in these plants. However the ‘flat’ ROL profiles along the lengths of the roots of all genotypes raised in deoxygenated solution suggest a partial barrier to ROL in the basal regions of these roots. The partial barrier resulted in rates of ROL from the most basal positions measured for adventitious roots raised in stagnant solution being similar to the rates, or in the case of L. elongatum, being half the rates from the same positions for roots of plants raised in aerated solution (Fig. 1), even though the porosity of the roots from the stagnant treatment were 1.3–5.3-fold higher. Nevertheless, ROL from an adventitious root of wheat (which has a partial barrier to ROL) to waterlogged soil can consume c. 20% of the O2 entering a root (Thomson et al., 1992), and this loss of O2 was suggested to contribute to poor adventitious root growth and intolerance of wheat to waterlogged soil (Thomson et al., 1992). In comparison waterlogging-tolerant rice not only has a larger volume of aerenchyma, but it also has a strong barrier to ROL in basal regions of its adventitious roots and therefore deeper root penetration into waterlogged soil (Armstrong, 1971; Thomson et al., 1992; Colmer et al., 1998).
In roots longer than the length that can be sustained by internal O2 diffusion in aerenchyma, only the basal tissues will receive adequate aeration and survival of the more apical regions may therefore be dependant upon fermentative metabolism. There was no statistical difference in the rate of ethanol production by excised, hypoxically pretreated, 10 mm adventitious root tips of the three genotypes (0.8–1.1 nmol g−1 f. wt s−1). These values were however, greater than those reported for 40 mm root tips of five wetland plant species immediately extracted from waterlogged soil in a wetland habitat which ranged from 0.23 nmol g−1 f. wt s−1 for Poa trivalis to 0.74 nmol g−1 f. wt s−1 for Filipendula ulmaria (Smith et al., 1986). Partial aeration of the root tips via internal O2 diffusion before excision and the inclusion of subapical regions of the root, may explain the lower rates of ethanol production reported by Smith et al. (1986). Root tips treated in a similar experimental fashion as in the present study (hypoxically pretreated, incubated with exogenous sugar), such as 10 mm maize root tips; 2.5 nmol g−1 f. wt s−1 (Hole et al., 1992) and 5 mm wheat root tips; 1.7–2.7 nmol g−1 f. wt s−1 (Waters et al., 1991b) gave faster ethanol production rates than were measured for the adventitious root tips of wheat and L. elongatum in the present study.
Evaluations of the L. elongatum × ‘Chinese Spring’ chromosome addition lines in stagnant solution showed that DA4E and the amphiploid displayed a slightly better tolerance than ‘Chinese Spring’ to growth in stagnant solution in terms of adventitious root d. wt and also, in the case of DA4E, shoot RGR. However, the small, but significant, differences in tolerance to growth in deoxygenated stagnant solution were not reproduced in waterlogged soil. Soil waterlogging is presumably a more severe stress than deoxygenated stagnant solution, which is reflected by the greater reductions in plant growth observed in waterlogged soil (Tables 3 and 6). This difference may reflect the accumulation of potentially phytotoxic reduced minerals and high rates of radial O2 loss from the roots that occur in waterlogged soil but not, or to a much lesser extent, in deoxygenated solution culture (Trought & Drew, 1980; Thomson et al., 1992; Watkin et al., 1998).
The present study showed that the amphiploid had a greater maximum root length than ‘Chinese Spring’ when plants were grown in deoxygenated stagnant solution. The longer roots were associated with a slightly higher porosity and slightly lower rate of O2 consumption in the amphiploid as compared with ‘Chinese Spring’, which together would have enhanced internal O2 diffusion to the apex (Table 2). Nevertheless, comparisons of the growth of ‘Chinese Spring’ with the amphiploid and chromosome addition lines in stagnant solution or in waterlogged soil showed small, or no significant differences, in tolerance to root zone O2 deficiency or to soil waterlogging. In an earlier study, the maximum lengths of adventitious roots in waterlogged soil were longer for the L. elongatum × wheat amphiploid and some of the addition lines, when compared with ‘Chinese Spring’ (Taeb et al., 1993). Maximum root lengths in an anaerobic medium are determined by traits (porosity, barriers to ROL, respiration rates) affecting internal O2 supply to the root apex (Armstrong, 1979). Unfortunately these traits were not evaluated in the plants grown by Taeb et al. (1993) so that the cause for the discrepancy between our findings can not be evaluated further. Nevertheless, the earlier view that the Lophopyrum× wheat lines showed substantial improvements in waterlogging tolerance (Taeb et al., 1993) was clearly not supported by the results of the present experiments.
We acknowledge the Grains Research and Development Corporation for their funding of this project and a student scholarship for MPM. We are grateful to Prof. Jan Dvořák for his generous donation of seeds for these experiments and to Prof. Roland von Bothmer for his assistance with the cytogenetic characterization of the lines.