Annals of Botany 90: 139-147, 2002
© 2002 Annals of Botany Company
A Simple Model of Feedback Regulation for Nitrate Uptake and N2 Fixation in Contrasting Phenotypes of White Clover
1 INRA, Unité d‘Agronomie, 234, Avenue du Brézet, F-63000 Clermont-Ferrand, France and 2 Plant Breeding Department, Institute of Grassland and Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, Wales, UK
* For correspondence. Fax +44 (0)1970 828257, e-mail frank.minchin{at}bbsrc.ac.uk
Received: 17 January 2002; Returned for revision: 6 March 2002; Accepted: 15 April 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
A simple three equation model is proposed for the feedback regulation of nitrate uptake and N2 fixation, based on the concentration of the organic N substrate pool within the plant and two parameters denoting the N substrate concentrations at which half-maximal inhibition occurs. This model simulated three contrasting phenotypes of white clover (Trifolium repens L.) inbred lines with (1) normal rates of nitrate uptake and N2 fixation (NNU); (2) low rates of nitrate uptake (LNU); and (3) very low rates of N2 fixation (VLF). The LNU phenotype was simulated by a decrease in the value of the inhibition parameter for nitrate uptake and the VLF phenotype was simulated by a decrease in the value of the N2 fixation inhibition parameter. The model was tested against nitrate uptake data obtained from white clover plants growing in flowing nutrient culture. There was an accurate prediction of the increase in nitrate uptake caused by N2 fixation activity of the NNU and LNU inbred lines being interrupted by a switch in gas phase from air to Ar : O2. The model was also tested against data for nitrate uptake, N2 fixation and %N from fixation for the three inbred clover lines grown in flowing nutrient culture at 0, 5 or 20 mmol m–3 NO3–. Again there was accurate prediction of nitrate uptake, although simulated values for N2 fixation were more variable. The simple model has potential use as a sub-routine in larger models of legume growth under field conditions.
Key words: Trifolium repens L., white clover, nitrate uptake, nitrogen fixation, phenotypes, regulation, models.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
It has been hypothesized that the nitrogen demand of the whole plant exerts a control on both nitrate uptake (Johnson, 1985; Imsande and Touraine, 1994) and nitrogen fixation (Hartwig, 1998; Serraj et al., 1999) through a feedback regulation based on the size of the soluble organic N pool within the plant. Although the exact nature of this feedback regulation is still uncertain, it is likely to utilize phloem-translocated amino acids as signal molecules [viz. Tillard et al. (1998) and Geßler et al. (1998) for nitrate uptake, and Neo and Layzell (1997) for N2 fixation]. However, it is unclear as to whether both forms of feedback regulation can coexist in nodulated legumes, where N acquisition through nitrate uptake is negatively correlated with N2 fixation activity (e.g. Macduff et al., 1996).
Indeed, parallel regulation of these two forms of N acquisition has not been included in most models of legume growth or grass/legume competition. For example, in the Hurley Pasture model (Thornley, 1998), a negative feedback regulation of inorganic N uptake was assumed, based on internal N levels, but symbiotic N2 fixation was not explicitly described. In an earlier model of the growth of grass–clover mixtures (Thornley et al., 1995), it was supposed that N2 fixation of the legume was insensitive to soil N, and N uptake rate was not affected by internal N. Then again, when modelling the coexistence of grass and clover, Schwinning and Parsons (1996) included a negative response of fixation to increasing soil mineral N and a positive response of N uptake. However, neither of these regulatory sub-models were based on the size of internal N pools. Recently, a negative feedback of substrate N concentration was assumed for both N2 fixation and N uptake in a phenomenological sub-model of grass–legume dynamics (Thornley, 2001). However, rather than a pure legume, a mixed sward was simulated in which a target legume content was assumed to depend on the carbon/nitrogen ratio in the plant substrate carbon and nitrogen pools.
Whilst these models have been very useful in understanding some of the complexity of grass/legume mixtures, a more detailed mechanistic approach may be required to improve the predictive accuracy of pasture models. Thus, in a recent model of the growth of pure and mixed grasses and legumes (Soussana and Oliveira Machado, 2000), both inorganic N uptake and N2 fixation activities of the simulated legume were assumed to be down-regulated according to Michaelis–Menten kinetics whenever the internal soluble organic N pool (substrate N) concentration increased. The sensitivity of symbiotic N2 fixation to substrate N content was then set to be substantially greater than that of inorganic N uptake. This difference in sensitivities allowed for the modelling of parallel regulation systems within the same plant.
However, there is limited experimental evidence regarding the parallel regulation of N2 fixation and nitrate uptake by legumes. Early work by Wych and Rains (1978) reported nitrate uptake by non-nodulated soybean to be more than twice that of nodulated plants, whilst more recent work of Kage (1995) and Soussana and Faurie (1998) suggested that the presence of active nodules had little effect on the nitrate uptake capacity of faba beans and white clover, respectively.
In the present work, the inter-relationship between N2 fixation and nitrate uptake regulation in legumes is addressed by measuring both activities in three highly inbred self-fertile lines of white clover. Two of these lines showed reduced rates of either nitrate uptake or N2 fixation (Michaelson-Yeates et al., 1998), possibly reflecting enhanced sensitivity to feedback regulation, whilst the third line, with normal rates of nitrate uptake and N2 fixation, acted as a control. For the low and normal nitrate uptake lines, the effect of a suppression of N intake from N2 fixation was observed by measuring nitrate uptake of roots supplied with an Ar:O2 atmosphere. All the experi mental results were then compared with simulations produced by a simple model of parallel feedback regulation derived from equations used by Soussana and Oliveira Machado (2000).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
Plant material and culture
The origin of the self-fertile inbred lines of white clover (Trifolium repens L.) has been described previously (Michaelson-Yeates et al., 1997, 1998; Abberton et al., 1998). In this study, the lines are designated as having normal nitrate uptake (NNU), low nitrate uptake (LNU) and very low fixation (VLF).
The main experiment involved plants of the NNU and LNU lines, which were raised from seed in potting compost and transplanted into flowing nutrient culture units (two units per line, each holding 24 plants). These provided automatic control and measurement of K+ and NO3– uptake and pH (Clement et al., 1974; Hatch et al., 1986). Each of the culture units contained 200 l of nutrient solution continuously recirculating through 24 x 1-l culture vessels each containing one plant. The units were situated in a glasshouse in which the air temperature was controlled at 20 ± 2 °C and the solution temperature at 20 ± 1 °C.
Preliminary experiments showed that for plants grown during the period December to February, a rapid exposure to high levels of artificial light produced phytotoxic symptoms (whitening of leaves) in these inbred clover lines. Therefore, a culture protocol was adopted that involved a much slower introduction of the artificial lighting regime. Thus, after transplanting, the 36-d-old seedlings initially received natural light for a further 13 d, after which supplementary lighting was introduced provided by a 400 W HPI-T lamp (Philips, Eindhoven, The Netherlands) over each culture unit [250 ± 25 µmol m–2 s–1 photosynthetically active radiation (PAR)], between 1000–1500 h. This supplementary illumination was increased incrementally over the following 24 d so as to replace natural light entirely with illumination (0800–1800 h) provided by a 400 W HPI-T and a 400 W SON-T lamp (Philips) over each culture unit (550 ± 50 µmol m–2 s–1 PAR).
Seedlings were initially transplanted into a nutrient solution containing 100 mmol NO3– m–3, which they were allowed to deplete for the next 37 d. After this time, plants were given fresh nutrient solution without N to allow for nodulation over a 20 d period. At the same time, each culture unit was inoculated with a mixture of five strains of Rhizobium trifolii (RCR 502, RCR 505, RCR 509, RCR 511 and RCR 515) known to be effective in symbiotic N2 fixation with T. repens. Nodules were visible on the root systems of inoculated plants after a further 10 d.
Experimental design
On day 93 after sowing, 24 plants of each line (NNU and LNU) were selected on the basis of uniform size and transferred to eight flowing nutrient ‘mini-culture units’ (six plants per unit) that were similar in design and operation to the main culture units, except that each contained 20 l of nutrient solution recirculating through six culture vessels. To expose the nodulated area of the root system to specific gas treatments, it was necessary to lower the liquid level in the culture vessels to provide an air space of approx. 0·1 l above the surface of the nutrient solution. This required moving the plants into new culture vessels in which the liquid overflow holes, which determine the level of liquid, were set at 7 cm from the top instead of the normal 3 cm.
Plants in these mini-tanks were subjected to atmospheres of either air (–Ar) or 79 % argon, 21 % O2 (+Ar). The plants designated for Ar : O2 exposure were sealed into the culture vessels, after which the composition of the nutrient solutions was controlled automatically and all units were subjected to a regime of very low NO3– supply, maintained at 5 ± 1 mmol m–3 NO3– by automatic delivery of Ca(NO3)2 on a measurement and addition cycle of 28 min. This concentration of nitrate was chosen to be high enough for accurate measurements of nitrate uptake but low enough to allow for continued nodule development and N2 fixation. Prior to the commencement of the Ar : O2 treatment, all eight mini-culture units were aerated by continuous bubbling of air, at a flow rate of 0·5 l min–1, through a sintered block located in the centre of the tank and supplied from an air pump. This aeration continued throughout the treatment period for the four mini-tanks (two for each plant line) that were not exposed to Ar : O2.
Plant sealing and Ar : O2 feeding
Following transfer into the mini-tanks, plants designated for Ar : O2 feeding (two units of each line) were sealed into their culture vessels. Sealing of the plant to the large central hole of the lid involved making a ‘cup’ of plastic sealing compound around the top of the tap root and filling the well of this cup with Dow Corning RTV 3110 silastic rubber compound (Dow Corning, Midland, MI, USA). Inlet and outlet air lines were sealed into smaller holes in the lid, using rubber sleeves, and the lid assemblies were then sealed onto the base of the containers using a non-phytotoxic silicon-based glazing sealant (Dow Corning).
Pumping a gas stream into the sealed gas space at the top of the flowing-nutrient culture vessels created a positive pressure that forced down the level of liquid in the vessels and allowed gas to escape out of the liquid overflow holes. Therefore, it was necessary to draw gas out of the gas space, thus creating a slight negative pressure which raised the liquid level and prevented gas loss from the overflow holes. This gas withdrawal was provided by a bank of air-pumps, whilst paired flow-meters situated in the inlet lines to the vessels and the outline lines of the pumps provided a check for leakage within the system.
The sealed vessels initially received air, at a flow rate of 0·2 l min–1. For Ar : O2 feeding, a gas stream of 79 % Ar (32·86 mol m–3) and 21 % O2 (8·74 mol m–3) was generated from gas cylinders using mass-flow controllers (ASM, Bilthoven, The Netherlands) and a KE25 electrochemical O2 sensor (Envin scientific products, Aylburton, UK) and fed into a manifold from which gas was drawn into each of the sealed culture vessels. A second manifold system supplied Ar : O2, at a flow rate of 0·5 l min–1, to the central aeration blocks for the four Ar : O2 mini-tank units. To ensure equilibration of pressure, an excess supply of Ar : O2 gas mixture was generated from the mass-flow controllers and the volume not required for the containers was bled to waste.
H2 production and respiration measurements
To provide a manageable number of gas lines for analysis, the 24 sealed culture vessels were linked together (i.e. the outflow from one culture vessel provided the inflow to the next vessel) to provide two groups of three linked vessels for each mini-tank. This provided a total of four replicate gas lines (two groups in two mini-culture units) for the LNU and NNU plants. The gas lines passed through pumps and then to computer-controlled solenoids which switched in turn every 7·5 min to complete a cycle of all eight gas lines every 60 min.
From the solenoids, the gas line for analysis passed through an electrochemical H2 detector (City Technology Ltd, Portsmouth, UK; Witty and Minchin, 1998) and then to an infrared gas analyser (ADC, Hoddesdon, UK) for determination of CO2 levels. The output from these was recorded with both a chart recorder and a computer, using the Labtec Notebook software package (Laboratory Technology Corporation, Wilmington, DE, USA).
The sealing operations induced a marked increase in nitrate uptake rates during the second and third day after sealing, with rates returning to the level of unsealed plants by the fifth day. The cause of this activity ‘spike’ is not known, but the return to the level of unsealed plants indicates that it was a transient perturbation. Plants in the four mini-tanks supplied with air throughout the experimental period were not sealed into their containers, and provided a control for determining the end of the sealing-induced perturbation period. This was taken as day 0 of the treatment period, and measurements of H2 and CO2 production commenced for the 24 Ar : O2 plants. For the first 2 d of measurements these plants continued to receive air, from which CO2 was removed with soda-lime scrubbers, and on day 2 the plants were switched onto Ar : O2, which they then received for the next 11 d. Data for the H2 production measurements are recorded as day 1 (for the 24 h measurement period day 0 to day 1) to day 13 (for the 24 h measurement period day 12 to day 13).
Uptake measurements and plant harvests
Daily net uptake of NO3– by plants in each mini-tank was given by the quantities of NO3– automatically delivered to maintain the set-point concentration of 5 mmol m–3 NO3–. Each unit was topped up at 0900 h each day with 5 l of nutrient solution of identical composition, to compensate for the volume of solution that was sampled for analysis and run to waste.
To estimate Vmax for NO3– uptake, the relationship between net uptake rate of NO3– and external concentration of NO3– for both inbred lines was measured at the start (run 1) and end (run 2) of the Ar : O2 exposure period by a depletion technique (Bakken et al., 1997). For run 1, five plants of each inbred line were selected randomly from those remaining in the main culture units. For run 2, three plants of each inbred line/treatment combination (LNU + air, LNU + Ar : O2, NNU + air, NNU + Ar : O2) were selected at random from one of each pair of replicate mini-tanks. The plants were transferred (0900 h) into separate polyethylene beakers with 1 l of aerated nutrient solution of the same composition as that in the mini-culture units, but with an increase to 150 mmol m–3 NO3–, and the addition of 10 mol m–3 MES buffer (pH 6·0). Samples (3 cm3) of uptake solution were collected every 10–20 min until NO3– was fully depleted from the solutions (5–6 h). Solution samples were analysed for NO3–, and the Michaelis–Menten parameters Vmax and Km calculated as described by Bakken et al. (1997). However, the goodness of fit of the Michaelis–Menten model was generally insufficient to give robust estimates of Km.
At the start of the Ar : O2 exposure period (day 2 of the treatment period), the supply of NO3– to all mini-tanks was replaced by a 15NO3– source (labelled at 5·0 atom %15N). At the end of the depletion runs all plants were harvested and separated into shoots and roots. A sub-sample of roots was then taken for nodule harvesting. All plant fractions were freeze-dried, weighed and ground to pass through a 0·5-mm sieve. Total N and 15N contents were measured by a continuous flow isotope mass spectrometer (Twenty-twenty; Europa Scientific Ltd, Crewe, UK) linked to a C/N analyser (Roboprep CN; Europa Scientific Ltd).
Following termination of the Ar : O2 treatment, the NO3– supply to the remaining air-treated plants was increased from a constant concentration of 5 mmol m–3 NO3– to one of 150 mmol m–3 NO3–. Net uptake of NO3– was subsequently measured automatically over a 44 h period.
Statistical analyses
Data for Tables 1 and 2 were analysed by ANOVA with one factor (genotype) for the initial harvest and two factors (genotype and argon) for the final harvest. Figures 1 and 2 were analysed by ANOVA for each day and a standard error of difference (s.e.d.) value was plotted. For Fig. 3, a single s.e.d. of means for genotype/treatment was plotted as ANOVA showed no significant changes with time.
|
|
|
Model description
The model predicts NO3– uptake and N2 fixation rates by legumes according to two variables: (1) the external concentration of NO3–; and (2) the internal concentration of ‘substrate N’, taken as the concentration of soluble organic N compounds in the plant that can be ‘sensed’ by the roots and nodules. The model has three equations and requires six parameters.
Equation (1) calculates the rate of NO3– uptake (Vupt, g N g–1 d. wt d–1) as the product of a Michaelis–Menten function of the NO3– concentration at the root surface ([NO3–]) and a potential value for NO3– uptake (Vupt-pot, g N g–1 d. wt d–1) which is down-regulated by the internal concentration of substrate N (Nsub, g N g–1 plant d. wt). Equation (2) assumes that the N2 fixation rate per unit root d. wt (Vfix, g N g–1 d. wt d–1) is also down-regulated from a potential value (Vfix-pot, g N g–1 d. wt d–1) by the internal N substrate concentration.
In eqns (1) and (2), Ku and Kf are the respective Nsub concentrations which halve the maximum rates of NO3– uptake and N2 fixation. In eqn (1), Km is the value of [NO3–] at which Vupt is half of the maximum nitrate uptake rate (Vupt-max).
As no estimate of Nsub was made in the present study, a third equation was added to predict Nsub from the measured plant N content (N%):
Nsub = (N% – N%min)/100(3)
where N%min is an estimate of the minimum plant N content for that phenotype.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
Plant growth and nodule development
Prior to exposure to Ar : O2 (day 2 of the treatment period), the inbred lines NNU and LNU had similar shoot and total plant dry weights, but LNU plants had smaller root and nodule weights (Table 1). During the treatment period, dry matter production was highest in the NNU line, irrespective of exposure to Ar : O2. However, the Ar : O2 treatment reduced the growth rate of both lines relative to control (air-treated) plants.
Nodule number and dry weight per plant increased substantially during the treatment period in both lines, although LNU plants had fewer nodules than NNU plants (Table 1). Exposure to Ar : O2 for 11 d had no significant effect on nodule development. Percentage N and total N content per plant were lower for the LNU plants at the beginning and end of the exposure period (Table 1), but increased substantially for both lines during this time. Exposure to Ar : O2 decreased %N and total N content for both lines by the end of the treatment. Moreover, the atom % 15N levels were higher in argon-treated plants.
H2 production
Nodules of plants of both lines, in air, produced similar amounts of H2 on days 1 and 2 of the treatment period and, following exposure to Ar : O2, H2 production increased by factors of 2 and 2·5, respectively, for LNU and NNU plants (Fig. 1). It remained at this level until day 4, then declined during the rest of the exposure period. Total root respiration (CO2 production) was very similar for both lines and showed no response to Ar : O2 exposure (data not shown).
Nitrate uptake
On day 1, following the start of H2 production measurements in air, the net uptake rate of NO3– by the LNU line was half that of the NNU line, on a per plant basis. Generally, the difference between the lines grew throughout the treatment period, irrespective of Ar treatment (Fig. 2). Exposure to Ar : O2 did not affect the increase in NO3– uptake by the NNU line until day 9. After this time (7 d after commencement of Ar : O2 exposure), uptake by +Ar plants increased faster than that by control plants. In contrast, NO3– uptake by +Ar and control plants of the LNU line remained almost constant until day 6, after which time uptake by +Ar plants increased steadily whilst that of control plants remained constant.
|
Vmax for NO3– uptake
Immediately prior to Ar : O2 treatment, Vmax for net uptake of NO3–, calculated from the depletion run data, was approx. three times higher for the NNU line compared with the LNU line (Table 2). The Vmax for control plants of both lines increased during the treatment period, but the value for –Ar NNU plants was still more than twice that for –Ar LNU plants by day 13. Exposure to Ar : O2 resulted in a sig nificant increase in the Vmax of both lines, by a factor of 1·3 for the NNU line and 4·2 for the LNU line.
|
At the end of the Ar : O2 exposure period, the air-treated plants were supplied with 150 mmol m–3 NO3– for a 44 h period (see Materials and Methods). The rate of NO3– uptake from this concentration was not significantly different from the apparent Vmax of air-treated NNU and LNU plants supplied with 5 mmol m–3 NO3– (Table 2). This suggests that under these steady-state conditions, NO3– uptake by control plants was not kinetically limited at a concentration of 5 mmol m–3 NO3–.
N2 fixation by the very low fixation (VLF) line
Measurements of H2 production by plants of the NNU and VLF lines, grown and sealed as for the Ar : O2 experiment (see Materials and Methods), were made over a 22 h period (Fig. 3). The rate of H2 production in air for VLF plants grown in flowing nutrient solution containing no NO3– (–N) was markedly lower than that for the comparable (–N) NNU plants, but similar to the rate of NNU plants supplied with 20 mmol m–3 NO3– (+N) for 21 d (Fig. 3). The rate of H2 production by +N VLF line plants was too low to be detected. Mean hourly rates of H2 production over 22 h were 10·02 (–N, NNU), 1·56 (+N, NNU) and 1·03 (–N, VLF) µmol per plant. Total nodule dry weights were 174, 87 and 26 mg per plant, respectively, giving specific activities of 57·7 (–N, NNU), 17·8 (+N, NNU) and 39·0 (–N, VLF) µmol H2 g1 d. wt h–1. However, the nitrogenase activity of the –N VLF plants was highly variable. Of the four plants selected for H2 production measurement, two gave activities below the level of accurate detection. This variation was confirmed using the acetylene reduction assay (ARA) in closed assay vessels as a ± determination of nitrogenase activity over a 5-week period. In all cases, zero or very low ARA activities were correlated with very low nodule weights (data not shown).
In a further experiment, H2 production by –N, VLF line plants was monitored during a change of the flowing gas phase from air to 79 % Ar, 21 % O2 (see Materials and Methods). This produced an initial doubling of H2 production, followed by a slow 30 % decline over the subsequent 2·5 h. The data were used to calculate a mean electron allocation coefficient (EAC) to N2 of 0·5 for the VLF line. This compares with a value of 0·62 for both the NNU and LNU lines, calculated from a preliminary experiment.
Calibration of model parameters and simulation results
Vupt-pot (0·03 g N g–1 root d. wt d–1) (Table 3) was fixed as the highest uptake rate measured during the depletion runs: the Vmax of LNU plants exposed to Ar : O2 (Table 2). Km (6 mmol m–3) was derived as the mean of all depletion runs in this study, plus previous (unpublished) runs with the same inbred lines. These values were used to calculate Ku for the NNU and LNU lines (Table 3) by fitting predicted nitrate uptake rates (Vupt) to measured uptake rates from 5 and 150 mmol m–3 NO3– on days 2 and 13 (Fig. 2; Table 2), using a non-linear regression procedure (Marquardt algorithm Statgraphics v5; Manugistics, San Diego, CA, USA). This procedure also required a value for the N substrate (reduced soluble N) pool, which could not be measured directly. It was initially estimated as a fixed proportion of total plant %N, but a much better fit was obtained when Nsub was calculated from plant %N data [see eqn (3), Materials and Methods] by assuming that Nsub tends towards 0 when the plant N concentration reaches a minimum value (N%min, Table 3). This minimum was then set at 3 for the NNU and LNU lines and 2 for the VLF line, as VLF plants displayed average %N values below 3. To obtain the best fit for the non-linear regression procedure (r2 = 0·64, n = 12 measurements, P < 0·01), it was necessary to use different Ku values for the two inbred lines, with the value for the LNU line (0·0013 ± 0·0003 g N substrate g–1 d. wt) being five times smaller than that for the NNU line (0·0065 ± 0·0015 g N substrate g–1 d. wt). The relationship between the predicted and measured values of Vupt is shown in Fig. 4.
|
|
The parameters for N2 fixation (Vfix-pot and Kf) could not be measured directly, and estimates had to be made by non-linear regression. Thus, Vfix-pot (0·04 g N g–1 root d. wt d–1, Table 3) was calculated as 1·5 times the highest measured rate of fixation activity (LNU plants in air; Table 2). To obtain the best fit for the non-linear regression procedure (r2 = 0·60, n = 10, P < 0·01), it was necessary to use different Kf values for the inbred lines, with the value for the LNU and NNU lines (0·0040 ± 0·0012) being ten times higher than that for the VLF line (0·0004 ± 0·0001).
The effects of Ar : O2 exposure on nitrate uptake were simulated (Table 4) by using eqns (1) and (3) of the model, with N substrate concentrations calculated from the measured values of total %N (Table 1). This predicts a 123 % increase in nitrate uptake by LNU plants, compared with a measured increase of 137 %. However, there was less agreement for NNU plants, with a predicted increase of 44 % and a measured increase of 11 %.
|
Further simulations were performed for NNU, LNU and VLF plants in air, using external concentrations of 0, 5 and 20 mmol m–3 NO3–, with values of Nsub calculated from measured values of %N. These predictions were compared with measured rates of nitrate uptake (Vupt), N2 fixation (Vfix) and %N from fixation (%fix), where %fix was calculated as 100 Vfix/(Vfix + Vupt). Data from Michaelson-Yeates et al. (1998), averaged over the treatment period, were used for plants with [NO3–] at 0 or 20 mmol m–3, together with data for the control NNU and LNU plants ([NO3–] = 5 mmol m–3) used in the present study (Fig. 5). There was very close agreement, to within a few per cent, between measured and predicted rates of NO3– uptake by NNU, LNU and VLF plants (n = eight measurements, r2 = 0·97). However, simulated values for Vfix were more variable (r2 = 0·60), primarily because the fixation rates by NNU and LNU plants at 5 mmol m–3 NO3– (calculated as total N increment minus nitrate N uptake) were higher than the rates for plants grown at 0 or 20 mmol m–3 NO3– in the previous study. Despite the variation between measured and predicted values of Vupt and Vfix, they showed a common trend in terms of %N derived from N2 fixation (r2 = 0·95), although there was a bias for VLF plants at 20 mmol m–3 NO3– as the %N from fixation was predicted to be positive by the model, whereas no detectable fixation could be measured.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
Physiological measurements
The coupling of a flow-through gas analysis system with an automated system of flowing solution culture is a novel means of obtaining continuous non-destructive measurements of both N2 fixation (as H2 production) and net uptake of NO3–. This approach has great potential for investigating the nitrogen nutrition of growing legumes under a range of conditions and treatments.
Nevertheless, long-term exposure of white clover plants to Ar/O2 produced an anomalous result in the slow decline of H2 production (Fig. 1). In theory, H2 production under Ar : O2 should have initially increased, as electron allocation by nitrogenase shifted from NH3 to H2, and then declined to a new steady-state level due to the Ar-induced decline (Minchin et al., 1983). This level of H2 production should then have been maintained, or even increased, during the remainder of the Ar : O2 exposure period. The observed decline in H2 production, to below the levels shown for plants in air, almost certainly represents a decrease in overall nitrogenase activity. It has previously been shown that excessive closure of the O2 diffusion barrier of legume nodules can lead to nodule deterioration due to oxidative damage (Escuredo et al., 1996). Although the study of Escuredo et al. (1996) reported a rapid loss of nodule activity induced by nitrate exposure, it is possible that a prolonged exposure to Ar-induced closure of the O2 diffusion barrier could produce a slower build-up of oxidative damage. Indeed, in a preliminary experiment it was found that exposure to Ar : O2 for 15 d caused substantial nodule senescence.
Despite this anomalous result, plants of both inbred lines exposed to Ar : O2 showed a decrease in N content and an increase in atom %15N (Table 1), indicating that their overall N balance had been altered. Indeed, exposure to Ar : O2 caused significant increases in measured rates of nitrate uptake (Fig. 2) and in Vmax as determined from depletion runs (Table 2), especially in the LNU plants.
These results strongly suggest that nitrate uptake by nodulated white clover plants is regulated by the plant’s N demand. Indeed, the reciprocity between N2 fixation and nitrate uptake is probably best explained by the existence of a parallel N feedback regulation system. Thus, a decrease in one form of N acquisition, due either to external factors or excessive sensitivity to internal substrate N, would produce a decrease in substrate N and allow for up-regulation of the other N source. This hypothesis is supported by the up-regulation of nitrate uptake under Ar : O2, of N2 fixation in LNU plants and of nitrate uptake in VLF plants (Michaelson-Yeates et al., 1998).
Modelling
The three equation, six parameter model can be used to simulate both nitrate uptake and N2 fixation by normally growing white clover, and to predict the effects of genetically determined variation in the sensitivity of nitrate uptake (LNU plants) and N2 fixation (VLF plants) to N substrate concentration (variation in Ku or Kf inhibition constants). The LNU phenotype can be accurately simulated by a decrease in Ku, which results in nitrate uptake being strongly depressed at moderate levels of N substrate, whilst N2 fixation is predicted to be higher than for the NNU line due to lower N substrate concentrations (suggested by lower %N levels in Table 1). However, nitrate uptake by NNU and LNU lines is predicted to be similar under very low [NO3–] or Nsub. The phenotype of the VLF line can be explained in terms of a low value for Kf, which results in a marked depression of N2 fixation at low levels of N substrate which have little or no effect on the other inbred lines. This assumption is supported by the H2 production rate measured for VLF plants, which was similar to that of NNU plants whose nodule activity was substantially down-regulated by a 21 d exposure to nitrate (Fig. 3).
The predicted NO3– uptake rates for all three inbred lines show reasonably good agreement with measured rates, both in terms of trends and absolute values (Fig. 5), whereas the agreement is more variable in the case of N2 fixation. Nevertheless, both the measurements and predictions indicate higher rates of N2 fixation by the LNU line, compared with the NNU line, and a decrease in fixation with increasing external nitrate concentration for both lines. Thus, the simulated trends for N2 fixation are in reasonable agreement with the measured ones.
A key feature of the model is the assumption of different Ku and Kf values indexed to a common N substrate pool. At a physiological level, differences in Ku or Kf could relate to different rates of phloem unloading in the root and nodule compartments. They could also be determined by the concentrations of different classes of organic N compounds (e.g. the amino acid spectrum), or the ratios between compounds.
Incorporation of the simple model as a sub-routine
The simple model used here provides a reasonably good quasi-mechanistic simulation of phenotypic variation in N acquisition by white clover growing in flowing nutrient culture. However, it may have greater value as part of a larger model to simulate legume growth under field conditions (e.g. Soussana and Oliveira Machado, 2000). Indeed, incorporation of this six parameter model as a subroutine may overcome some of its limitations. The present model uses a minimum %N value as a ‘threshold factor’ in the calculation of N substrate concentration from total plant %N values and does not incorporate a direct link between external N levels and N substrate concentration. However, in larger models, N substrate concentration can be simulated from the effects of external N on the rates of N acquisition and the effects of N demand on N assimilation into protein synthesis (e.g. the Canopt model; Soussana and Oliveira Machado, 2000).
Nevertheless, there are some potential problems that may limit incorporation of the simple model. From the present simulations, the procedures for estimation of the nitrate uptake parameters (Vupt-pot and Ku) would appear to be reasonable for hydroponically grown plants. With field-grown plants, a similar nitrate depletion method has been used with excised roots (e.g. Léon et al., 1995). However, estimations of the N2 fixation parameters (Vfix-pot and Kf) are more problematical. At present, there are no procedures for measurement of Vfix-pot, and its estimation required an iterative process which used the model to obtain best fits against measured values of N2 fixation. Furthermore, in the present simulations, Vupt-pot and Vfix-pot are assumed to be fixed values, whilst they are likely to vary with time under field conditions. This is exemplified in this study by the apparent decline in the potential nodule activity (H2 evolution rate) of plants under Ar : O2 (Fig. 2), which would be best represented by a decline in Vfix-pot.
Then again, down-regulation from Vfix-pot will involve changes in both nodule mass and nodule activity, whilst, in the present model, nodule mass is assumed to be a fixed proportion of root mass. This is clearly not true in the case of VLF plants where the effect of N substrate is greater on nodule mass production than on the activity of the remaining nodules. A distinction between nodule and root mass would be a useful addition to a mechanistic model of legume growth, but would require accurate data for nodule production and senescence under field conditions.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
The simple six parameter model presented in this paper is an attempt at a mechanistic model of the feedback regulation of both nitrate uptake and N2 fixation within the same root system. The model has a good heuristic value and is not in contradiction with the available data. However, more stringent proofs are required for its validation under field conditions.
|
ACKNOWLEDGEMENTS |
|---|
The Unité d’Agronomie is funded by INRA, and IGER is funded by BBSRC. The support of an INRA/BBSRC programme grant is gratefully acknowledged.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSLITERATURE CITED |
|---|
-
Abberton MT, Michaelson-Yeates TPT, Macduff JH. 1998. Charac terization of novel inbred lines of white clover (Trifolium repens L.). I. Dynamics of plant growth and nodule development in flowing solution culture. Euphytica 103: 35–43.
Bakken AK, Macduff JH, Humphreys M, Raistrick N. 1997. A stay-green mutation of Lolium perenne affects NO3– uptake and translocation of N during prolonged N starvation. New Phytologist 135: 41–50.[CrossRef]
Clement CR, Hopper MJ, Canaway RJ, Jones LHP. 1974. A system for measuring the uptake of ions by plants from solutions of controlled composition. Journal of Experimental Botany 29: 453–464.
Escuredo PR, Minchin FR, Gogorcena Y, Iturbe-Ormaetxe I, Klucas RV, Becana M. 1996. Involvement of activated oxygen in nitrate-induced senescence of pea root nodules. Plant Physiology 110: 1187–1195.[Abstract]
Geßler A, Schultze M, Schrempp S, Rennenberg H. 1998. Interaction of phloem-translocated amino compounds with nitrate net uptake by the roots of beech (Fagus sylvatica) seedlings. Journal of Experimental Botany 49: 1529–1537.
Hartwig UA. 1998. The regulation of symbiotic N2 fixation: a conceptual model of N feedback from the ecosystem to the gene expression level. Perspectives in Plant Ecology, Evolution and Systematics 1: 92–120.[CrossRef]
Hatch DJ, Hopper MJ, Dhanoa MS. 1986. Measurement of ammonium ions in flowing solution culture and diurnal variation in uptake in Lolium perenne L. Journal of Experimental Botany 37: 589–596.
Imsande J, Touraine B. 1994. N demand and the regulation of nitrate uptake. Plant Physiology 105: 3–7.[Web of Science][Medline]
Johnson IR. 1985. A model of the partitioning of growth between the shoots and roots of vegetative plants. Annals of Botany 55: 421–431.
Kage H. 1995. Interaction of nitrate uptake and nitrogen fixation in faba beans. Plant and Soil 176: 189–196.[CrossRef]
Léon M, Lainé P, Ourry A, Boucaud J. 1995. Increased uptake of native soil nitrogen by roots of Lolium multiflorum Lam. after nitrogen fertilization is explained by a stimulation of the uptake process itself. Plant and Soil 173: 197–203. [CrossRef]
Macduff JH, Jarvis SC, Davidson IA. 1996. Inhibition of N2 fixation by white clover (Trifolium repens L.) at low concentrations of NO3– in flowing solution culture. Plant and Soil 180: 287–295.[CrossRef]
Michaelson-Yeates TPT, Macduff JH, Abberton MT, Raistrick N. 1998. Characterization of novel inbred lines of white clover (Trifolium repens L.) II. Variation in N2 fixation, NO3– uptake and their interactions. Euphytica 103: 45–54.
Michaelson-Yeates TPT, Marshall A, Abberton MT, Rhodes I. 1997. Self-compatibility and heterosis in white clover (Trifolium repens L.). Euphytica 94: 341–348.[CrossRef]
Minchin FR, Witty JF, Sheehy JE, Muller M. 1983. A major error in the acetylene reduction assay: decreases in nodular nitrogenase activity under assay conditions. Journal of Experimental Botany 34: 641–649.
Neo HH, Layzell DB. 1997. Phloem glutamine and the regulation of O2 diffusion in legume nodules. Plant Physiology 113: 259–267.[Abstract]
Schwinning S, Parsons AJ. 1996. Analysis of the coexistence mechanisms for grasses and legumes in grazing systems. Journal of Ecology 84: 799–813.[CrossRef]
Serraj R, Sinclair TR, Purcell LC. 1999. Symbiotic N2 fixation response to drought. Journal of Experimental Botany 50: 143–155.
Soussana JF, Faurie O. 1998. The regulation of N2 fixation and NO3– uptake in grass-clover mixtures. In: Elmerich C, Kondorosi A, Newton WE, eds. Biological nitrogen fixation for the 21st century. Dordrecht: Kluwer Academic Publishers, 662.
Soussana JF, Oliveira Machado A. 2000. Modelling the dynamics of temperate grasses and legumes in cut mixtures. In: Lemaire G, Hodgson J, Moraes C, Nabinger C, Carvalho PC de F, eds. Grassland ecophysiology and grazing ecology. Wallingford, UK: CABI International, 169–190.
Thornley JHM. 1998. Grassland dynamics: an ecosystem simulation model. Wallingford: CABI International.
Thornley JHM. 2001. Simulating grass-legume dynamics: a pheno menological submodel. Annals of Botany 88: 905–913.
Thornley JHM, Bergelson J, Parsons AJ. 1995. Coupled dynamics in a carbon-nitrogen model of a grass-legume pasture. Annals of Botany 75: 79–94.
Tillard P, Passama L, Gojon A. 1998. Are phloem amino acids involved in the shoot to root control of NO3– uptake in Ricinus communis plants? Journal of Experimental Botany 49: 1371–1379.
Witty JF, Minchin FR. 1998. Methods for the continuous measurement of O2 consumption and H2 production by nodulated legume root systems. Journal of Experimental Botany 49: 1041–1047.
Wych RD, Rains DW. 1978. Simultaneous measurement of nitrogen fixation estimated by acetylene-ethylene assay and nitrate absorption by soybeans. Plant Physiology 62: 443–448.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Z. YANG and D. J. MIDMORE A Model for the Circadian Oscillations in Expression and Activity of Nitrate Reductase in Higher Plants Ann. Bot., November 1, 2005; 96(6): 1019 - 1026. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||