Transpiration, Potassium Uptake and Flow in Tobacco as Affected by Nitrogen Forms and Nutrient Levels

Abstract

• Background and Aims Ammonium can result in toxicity symptoms in many plants when it is supplied as the sole source of N. In this work, influences of different nitrogen forms at two levels (2 and 15 mm N) on growth, water relations and uptake and flow of potassium were studied in plants of Nicotiana tabacum ‘K 326’.

• Methods Xylem sap from different leaves was collected from 106-d-old tobacco plants cultured in quartz sand by application of pressure to the root system. Whole-shoot transpiration for each of the treatments was measured on a daily basis by weight determination.

• Key Results Total replacement of by caused a substantial decrease in dry weight gain, even when plants grew under nutrient deficiency. Increasing nutrient concentration resulted in a greater net dry weight gain when nitrogen was supplied as or NH₄NO₃, but resulted in little change when nitrogen was supplied as . as the sole N-source also caused reduction in transpiration rate, changes in plant WUE (which depended on the nutrient levels) and a decrease in potassium uptake. However, the amount of xylem-transported potassium in the plants fed with was not reduced: it was 457 % or 596 % of the potassium currently taken up at low or high nutrient level, respectively, indicating a massive export from leaves and cycling of potassium in the phloem.

• Conclusions Ammonium reduces leaf stomatal conductance of tobacco plants. The flow and partitioning of potassium in tobacco plants can be changed, depending on the nitrogen forms and nutrient levels.

INTRODUCTION

Ammonium () and nitrate () are two main forms of nitrogen (N) that can be absorbed by higher plants from the natural environment. Because the two forms are fundamentally different in charge, they have quite different effects on the physiological and metabolic processes of plants (Haynes and Goh, 1978; Marschner, 1995). Although is an intermediate in many metabolic reactions, it can result in toxicity symptoms in many plants when it is supplied as the sole nitrogen source (Britto and Kronzucker, 2002).

One of the inhibitory effects of on plant growth is its influence on water relations. It has been reported that caused an increase in stomatal conductance in white clover (Høgh-Jensen and Schjoerring, 1997), and enhanced the transpiration rate in alfalfa (Khan et al., 1994) and tomatoes (Lugert et al., 2001). In both cases, leaf water and osmotic potential was lower than that of plants grown under conditions (Ashraf, 1999). However, in Phaseolus vulgaris, was found to decrease stomatal conductance and reduce root water uptake (Guo et al., 2002). Other research using Beta vulgaris indicated that the N-source had no effect on stomatal conductance, although use of did reduce water uptake in muskmelon and sugar beet (Raab and Terry, 1994; Adler et al., 1996), with a slight reduction in leaf conductance in castor beans (Peuke et al., 1994a). The dual-affinity nitrate transporter gene AtNRT1.1 (CHL1) is expressed not only in roots, but also in guard cells of Arabidopsis thaliana plants and functions as a transporter, and it has been demonstrated that CHL1 supports stomatal function in the presence of (Guo et al., 2003). The results imply that plays a role in stomatal opening, since in the absence of stomatal aperture and transpiration are reduced.

nutrition also causes a decrease in water use efficiency (WUE) in tomatoes (Lugert et al., 2001; Claussen, 2002), white clover (Høgh-Jensen and Schjoerring, 1997), Phaseolus vulgaris and Ricinus communis (Raven et al., 1992), but increases WUE in wheat (Morgan, 1986; Yin and Raven, 1998). It seems that the influence of on water relations varies depending on the plant species and the experimental conditions. Given these variable responses, it would be useful to determine within species how water relations behave when plants grow under various nutritional regimes and supplied with different nitrogen forms. Stable carbon isotope discrimination (δ¹³C) is a useful method to study plant photosynthesis (Farquhar et al., 1989). It is known that variations in the ¹³C/¹²C ratio of the organic carbon of plants are a result of genetic and environmental factors (Farquhar et al., 1989; Raven et al., 1992). N-sources, for example, can affect plant δ¹³C by influencing plant WUE and carbon assimilation (Yin and Raven, 1998). A common finding is that a lower WUE is correlated with a more negative δ¹³C value (Farquhar et al., 1989; Raven and Farquhar, 1990; Raven et al., 1992; Yin and Raven, 1997, 1998).

has been found to cause a strong inhibition in potassium uptake by plants (Allen and Raven, 1987). Potassium, as a macro-element in the cells of higher plants, exhibits high mobility. Its flow, however, is quite different depending on plant type (Armstrong and Kirkby, 1979; Jeschke et al., 1987; Jiang et al., 2001) and can be changed by stress. Under saline conditions, for example, plants have been found to have higher potassium cycling rates compared with plants grown under normal conditions. This phenomenon may contribute to the plant's ability to adjust to saline stress (Jeschke et al., 1987; Jeschke and Pate, 1991). Plants infected by a parasitic angiosperm (Hibberd et al., 1999) or that have been decapitated (Jiang et al., 2001) have also shown changes in potassium flow and cycling patterns within them. However, the extent to which translocation in the xylem and circulation of potassium is altered by the supply of N in different forms under varying nutrient conditions is not clear.

The present experiments were carried out to address these questions. The influence of different N-forms at two nutrient levels on the growth, water relations, and potassium uptake and flow in tobacco plants was investigated to better understand the underlying mechanisms of the inhibitory effects of on plant growth.

MATERIALS AND METHODS

Plant cultivation

Tobacco seeds (Nicotiana tabacum L. ‘K 326’) were germinated in a mixture consisting of 60 % (w/w) peat culture substrate, 20 % (w/w) ground maize stalk and 20 % (w/w) perlite, and grown in a seedbed in a naturally illuminated greenhouse for 65 d. The plants were then washed with tap water to remove all substrate from the roots, and transferred to 2·3 L pots (one plant per pot) containing quartz sand (0·25–0·5 mm in diameter). They were supplied initially with a half-strength nutrient solution as used by Walch-Liu et al. (2000). The solution consisted of (in mm for full strength): KH₂PO₄ 0·5, MgSO₄ 1·2, CaCl₂ 2·0, KNO₃ 2, H₃BO₃ 1·0 × 10⁻², MnSO₄ 5 × 10⁻⁴, ZnSO₄ 5 × 10⁻⁴, CuSO₄ 1 × 10⁻⁴, (NH₄)₆Mo₇O₂₄ 1 × 10⁻⁵, Fe-EDTA 1·5 × 10⁻². After 2 d a full strength solution was provided. The pots were watered every morning with an excess of the nutrient solution, a hole in the bottom of the pot allowing drainage. Plants were grown in a growth room with a 14-h photoperiod. The photosynthetically active radiation at the surface of the pots was 220–270 mmol m⁻² s⁻¹ provided by reflector sunlight dysprosium lamps (DDF 400, Nanjing, China). The first harvest was made 106 d after sowing and 41 d after transferring to constant environmental conditions; the second harvest was performed 10 d later.

Treatments and harvest procedures

For harvest, the plants were divided into seven groups of five plants, each of similar size and development. One group was used for the first harvest, the other six groups for the second harvest. On the day of the first harvest, the remaining six groups of plants were treated with either a low (as used by Walch-Liu et al., 2000) or high (Hoagland solution) concentration of nutrient solution combined with different forms of nitrogen (, and NH₄NO₃) as shown in Table 1.

Table 1.

Macronutrient composition of the solutions with low or high concentration of nutrients combined with different N-forms (in mm)

	Treatment
Composition	Low	Low	Low NH4NO3	High	High	High NH4NO3
KH2PO4	0·5	0·5	0·5	1	1	1
KNO3		2			5
NH4NO3			1			7·5
(NH4)2SO4	1			7·5
Ca(NO3)2					5
CaCl2	2·0	2·0	2·0	5		5
MgSO4	1·2	1·2	1·2	2	2	2
K2SO4	1		1	2·5		2·5

In addition, the media with low concentration of nutrients contained the same micronutrient as mentioned in the section ‘Plant cultivation’. The high concentration of nutrients contained the following micronutrients (mm): H₃BO₃ 4·6 × 10⁻², ZnSO₄.7H₂O 7·6 × 10⁻⁴, CuSO₄.5H₂O 3·2 × 10⁻⁴, (NH₄)₆MoO₂₄ 1·6 × 10⁻⁵, MnCl₂.4H₂O 9·0 × 10⁻³, FeEDTA 3·7 × 10⁻².

Plant leaves were numbered in ascending order, starting with the lowest mature leaf, which was designed as leaf 1. Smaller leaves, which had already senesced, were removed. At the first harvest, the youngest unfolded leaf was number 9, and the top part of the plant above this was incorporated into the upper leaves because of its small size. At harvest, the plants were separated into the lower, middle and upper stratum of three leaves each, and also into roots, stem and top. For the second harvest the top included new leaves that had grown out above leaf 9. Roots were washed free of sand with tapwater. The root and the three strata of leaves were divided into two lateral symmetrical parts: one was kept at −20 °C until analysis of tissue and contents, and the other was dried (70 °C) for determination of ions present in different organs with Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP) (Perkin Elmer 3300 DV, USA).

For the measurement of tissue and contents, samples were homogenized with distilled water and centrifuged. The residue was washed twice with distilled water by repeating the homogenization/centrifugation procedure. The combined supernatants were used for analysis of tissue and contents with a TRAACS-2000 autoanalyser (Bran+Luebbe, Germany).

Measurement of transpiration

Whole-shoot transpiration was measured on a daily basis by weighing five pots of each of the treatments at the beginning of the light period, after the daily addition of nutrient solution and draining, and at the end of the light period (08:00 h, 10:00 h and 22:00 h, respectively). The transpiration between 08:00 h and 10:00 h was calculated by averaging the amount of transpiration between 10:00 h and 22:00 h. Corrections were applied for water loss from pots without plants. In all cases, evaporation from the pots was minimized by covering the exposed surface with a layer of filter paper.

Collection of xylem sap

For collection of xylem sap, plants were grown in special pots in order to apply pressure to the root system (Seel and Jeschke, 1999), but were otherwise treated the same as the plants for harvest. The xylem sap collection procedure was essentially the same as described by Jeschke and Pate (1991). Samples were taken from leaf numbers 2, 5 and 8 (counting up from the base). In brief, approximately midway along the length of the leaf an incision was made into the midrib. The cut surface was washed carefully with distilled water and a Teflon tube attached. Pressure was applied to both the sand substrate and the roots of the treated plants. After slowly applying pressure, xylem sap started to exude from the midrib after reaching a balancing pressure (Seel and Jeschke, 1999), and sap was collected 50 kPa above this pressure. The first exudate was discarded to avoid contamination from damaged cells. Xylem sap was also collected from the stem base of the pressurized plants, either from an incision into the stem (Hibberd et al., 1999) or by inserting a syringe needle into the stem until it reached the xylem vessels. Xylem sap was kept on ice during collection and stored at −20 °C before analysis. After appropriate dilution, ions were analysed directly using ICP (PE 3300 DV, USA).

Measurement of carbon isotope discrimination (δ¹³C)

Leaves (including leaf veins and petioles) were dried at 70 °C and ground into a fine powder. They were then treated and analysed for carbon isotope composition using a mass spectrometer (Delta S, Finnigan MAT, Germany) according to the method described by Farquhar et al. (1989), with three internal replications. The ¹³C/¹²C ratios of the samples were expressed as δ¹³C relative to the isotope composition of the fossil belemnite standard from the Pee Dee Formation (PDB).

Estimation of the net flows of potassium through xylem and phloem in the whole plant

Net flows of potassium within plants were estimated using the method described by Armstrong and Kirkby (1979). Based on the assumption that nutrients were transported solely through xylem and phloem, Ca²⁺ could only be transported apically through the xylem and had no mobility in the phloem.

The net xylem flow of Ca²⁺, J_Ca,x, transported to various plant parts during a given period was the same of that of the net increment of Ca²⁺, ΔCa, in the same position during the same phase, where

The content of K⁺ and calcium had a stable ratio in the xylem during relatively stable growth conditions. The [K/Ca]_x ratio was obtained by determining the content of these two elements in the xylem sap collected.

For a given period, the net xylem flow of potassium, J_K,x, towards a site was calculated from the net increment of Ca²⁺, ΔCa, and the ratio of potassium and calcium, [K/Ca]_x, in the xylem by:

The amount of potassium exported through the phloem, J_K,p, was equal to the difference between the measured potassium increment, ΔK, in each organ and the net xylem import according to:

A positive difference indicated net phloem import, while a negative difference implied net phloem export from an organ. Working progressively along the plant from the root to the leaves and top, the net flows of potassium within the whole plant were obtained as shown in Figs 3 and 4. The differences between the quantities of potassium translocated in the phloem and in the xylem also allowed the estimation of transfer processes between these translocation streams.

Fig. 3.

Flow profiles for uptake, transport and utilization of K⁺ in tobacco plants supplied with a low nutrient concentration (2·5 mm K and 2 mm N) over a 10 d experimental period, starting 106 d after sowing. The values of K⁺ deposition and the statistical significance are given in Table 3. The width of arrows and the height of histograms are drawn in proportion to the net flows and deposition of K⁺. The numbers indicate the values of uptake, transport and utilization (mmol K⁺ per plant over the 10 d study period).

Fig. 4.

Flow profiles for uptake, transport and utilization of K⁺ in tobacco plants supplied with a high nutrient concentration (6 mm K and 15 mm N) over a 10 d experimental period, starting 106 d after sowing. The values of K⁺ deposition and the statistical significance are given in Table 3. The width of arrows and the height of histograms are drawn in proportion to the net flows and deposition of K⁺. The numbers indicate the values of uptake, transport and utilization (mmol K⁺ per plant over the 10 d study period).

Statistical treatment

Dry weight increments were obtained from the five replicates of each treatment at the first and second harvest. All further analyses were made with five individual samples for each organ. For statistical analysis of the data, SAS for Windows (version 6.12) was used (SAS, 1987). Differences between data in all tables were tested with ANOVA.

RESULTS

Plant growth and development

At low nutrient levels, all plants treated with the three forms of nitrogen showed symptoms of N deficiency, such as thin leaves and light-green lower leaf colour, and small plant size at the second harvest compared with those grown at higher N-levels. However, the growth of the plants was more inhibited than those of either NO₃- or NH₄NO₃-fed plants. The total net increment in dry weight of the plants was 23·5 % and 22·7 % lower in comparison with that of the NO₃- and NH₄NO₃-fed plants, respectively. Increasing nutrient concentration resulted in a greater net dry weight gain when nitrogen was supplied as or NH₄NO₃, but resulted in little change when nitrogen was supplied as . The total net dry weight gain of the plants was 38·8 % and 47·9 % lower than that of the NO₃- and NH₄NO₃-fed plants, respectively (Table 2). Under the and NH₄NO₃ treatments at both nutrient-levels, the plant top (including new leaves and upper leaves) contributed the most to the dry weight increment, while under the treatment, the upper and middle leaves were the main sites for assimilate deposition. In spite of the inhibited growth of the plants treated with different N-forms at low nutrient-level during the study period, their root net increments were not reduced in comparison with the respective treatments at high nutrient-level (Table 2), resulting in a decreased shoot/total dry weight ratio.

Table 2.

Initial values and net increments in dry weights of component organs and of whole tobacco plants under different treatments over a 10 d study period

		Net increase in DW (g organ−1)
		Low concentration					High concentration
Position	Initial DW (g organ−1)			NH4NO3			NH4NO3
Top + new leaves		0·53bc	0·95b	0·79bc	0·42c	2·05a	2·21a
Upper leaves	0·43 ± 0·05	0·77c	1·15b	1·16b	0·80c	1·49a	1·61a
Middle leaves	0·48 ± 0·08	0·81ab	0·79ab	0·88a	0·84ab	0·71b	0·90a
Lower leaves	0·43 ± 0·15	0·40a	0·44a	0·53a	0·53a	0·39a	0·51a
Stem	0·58 ± 0·15	0·64b	0·69ab	0·54b	0·62b	0·71ab	0·84a
Roots	0·38 ± 0·18	0·62bc	0·90abc	0·98ab	0·57c	0·82abc	1·18a
Whole plant	2·30 ± 0·52	3·77d	4·93c	4·88c	3·78d	6·18b	7·26a

Values in rows followed by the same letter are not significantly different (P > 0·05).

Total transpiration and leaf δ¹³C

When a high level of nutrients was supplied, the shoot transpiration of the NO₃- and NH₄NO₃-fed plants increased soon after the beginning of the treatment, remaining stable until the end of the experiment. Shoot transpiration declined during the same time for plants fed with (Fig. 1). At low nutrient level, the shoot transpiration of the NO₃- and NH₄NO₃-treated plants did not increase as much as that of the plants supplied with the same N-forms at higher nutrient level, but was still higher than that of the plants.

Fig. 1.

Time-course of whole-shoot transpiration of tobacco plants as dependent on N-forms and nutrient. Bars denote s.e. of the mean, n = 5.

At the end of the experiment, samples collected from the middle and upper leaf strata and top and new leaves of the - and NO₃-fed plants were used to compare their stable carbon isotopes δ¹³C. Tissue δ¹³C in the treatment was less negative (smaller absolute value) than that in the treatment at either nutrient level; especially in the top and new leaves under high nutrient level (Fig. 2).

Fig. 2.

Carbon isotope fractionation (δ¹³C, ‰) of leaf tissues of - and NO₃-fed plants at both low and high nutrient levels during a 10 d study period. L and H represent the low and high nutrient level, respectively. Bars denote s.e. of the mean, n = 5.

and contents in leaf tissues and roots

As shown in Table 3, no significant differences in and content was observed in the leaves and roots of the plants treated with three N-forms at low nutrient level. At high nutrient level, however, the content in the upper and middle leaves of the plants was substantially lower than that of the NO₃- and NH₄NO₃-treated plants, and could not be detected in lower leaves and roots. The content, on the other hand, was obviously higher in the upper leaves and roots of both ( and NH₄NO₃) plants than that of the NO₃-fed plants.

Table 3.

Contents ofandin different leaves and roots of tobacco plants grown under different nutrient concentrations and nitrogen forms over a 10 d study period (mmol organ⁻¹)

	Treatment
Position	Low	Low	Low NH4NO3	High	High	High NH4NO3
content
Upper leaves	2·65b	2·71b	2·37b	35·17a	4·34b	23·25a
Middle leaves	1·81b	1·29b	1·68b	16·41a	6·81ab	12·75ab
Lower leaves	1·81a	0·92a	2·01a	3·64a	1·30a	2·22a
Roots	12·55bc	4·52c	3·78c	21·90ab	6·18c	28·32a
content
Upper leaves	–	–	–	0·70c	231·87a	96·96b
Middle leaves	–	–	–	1·77b	58·88a	40·84a
Lower leaves	–	–	–	–	20·49a	21·97a
Roots	–	–	–	–	52·92a	24·06b

Values in rows followed by the same letter are not significantly different (P > 0·05). – indicates not detected.

Effect of N supply on contents and changes in potassium

The changes in net potassium increment as affected by different N-forms and nutrient levels were similar to, but clearly more dramatic than those in net dry weight gain. The total net uptake of potassium in plants was markedly smaller than that in the NO₃- or NH₄NO₃-fed plants at both nutrient levels (Table 4). The greatest increments of potassium were found in the rapidly growing organs, i.e. top and new leaves, and upper leaves of the NO₃- and NH₄NO₃-fed plants, and top and new leaves, and upper and middle leaves of the plants. Substantial increments were also found in the middle leaves, stem and roots of the NO₃- and NH₄NO₃-treated plants, while that in stem and roots of the plants was much less at both nutrient levels during the study period. Again, as with dry weight gain, the total potassium K⁺ contents in the NO₃- and NH₄NO₃-treated plants increased when potassium supply was elevated. The potassium content of the plants was relatively unchanged under the same conditions.

Table 4.

Initial values and net increments in K⁺ contents of component organs and of whole tobacco plants under different treatments over a 10 d study period

		Net K+ increment (mmol organ−1)
		Low concentration					High concentration
Position	Initial values (mmol organ−1)			NH4NO3			NH4NO3
Top+new leaves		0·52c	1·37b	1·08bc	0·46c	3·05a	3·37a
Upper leaves	0·49 ± 0·07	0·68c	1·56b	1·65b	0·81c	2·29a	2·45a
Middle leaves	0·53 ± 0·06	0·52c	0·88b	0·96b	0·60c	1·01b	1·36a
Lower leaves	0·39 ± 0·12	0·25d	0·41bc	0·49ab	0·33cd	0·50ab	0·58a
Stem	0·37 ± 0·06	0·18d	0·68b	0·49bc	0·25cd	1·02a	1·14a
Roots	0·17 ± 0·06	0·15b	0·88a	1·00a	0·11b	0·92a	1·06a
Whole plant	1·96 ± 0·21	2·30d	5·78c	5·67c	2·56d	8·81b	9·97a

Values in rows followed by the same letter are not significantly different (P > 0·05).

Water use efficiency of plants

The WUE is presented as net increase in plant dry weight per unit water used (g DW increase kg⁻¹ H₂O) during the experimental period. The WUE of the NO₃- and NH₄NO₃-fed plants was almost same under both nutrient levels, but that of the plants was the lowest and that of the plants was the highest among the treatments (Table 5).

Table 5.

Water use efficiency (g DW increase kg⁻¹ H₂O) in different treatments over a 10 d study period

	Low concentration					High concentration
			NH4NO3			NH4NO3
WUE	5·69	6·49	6·12	6·89	5·89	6·07

Estimation of net flows of potassium within tobacco plants

As shown in Figs 3 and 4, in spite of the inhibited K⁺ uptake in the plants, K⁺ transport in the xylem was still more than that in the NO₃-fed plants. The highest amount of the xylem-transported potassium was found in the NH₄NO₃-treated plants at both nutrient levels. The amount of the xylem-transported potassium was much more than accounted for by uptake, and was higher at high nutrient level than that at low nutrient level. Phloem retranslocation of potassium from shoot to root contributed to the xylem-transported potassium. The amount of the phloem-retranslocated potassium in the plants was four times more than that of the total uptake under high nutrient levels and 2·6 times more under low levels.

The phloem-recycled potassium from shoot to roots came mainly from leaves (Figs 3, 4). The proportion of potassium exported via the phloem to that imported via the xylem of a leaf depended on leaf age, nutrient levels and N-forms. In plants fed with a high level of NO₃-N, for example, it was 45 %, 70 % and 78 % in the upper, middle and lower leaves, respectively; while under high conditions, it was increased to 60 %, 92 % and 94 %, respectively. Also at low nutrient level, the value was 35 %, 68 % and 85 % respectively in the NO₃-fed plants, and 42 %, 89 % and 91 % in the plants.

DISCUSSION

Effect on plant growth

The reported symptoms of toxicity generally appeared with external concentrations above 0·1–0·5 mm (reviewed in Britto and Kronzucker, 2002). In spite of the inhibited growth under low nutrient levels (2 mm N), as represented by lower net dry weight gain (Table 2), lower shoot/total dry weight ratio and obvious N-deficient symptoms (photographs not shown), plant growth was further suppressed when was supplied as the sole N-source. Increasing nutrient concentration (Hoagland, 15 mm N) resulted in a greater net dry weight increment when nitrogen was supplied as or NH₄NO₃, but produced no change when nitrogen was supplied as (Table 2). These results could not simply be explained by toxicity. In the present study, the same amount of was found in the leaf tissues and roots of the plants grown under the low nutrient level with different N-forms. Furthermore, at the high nutrient level, a similar content was detected in the tissues of both - and NH₄NO₃-fed plants (Table 3): the latter were able to grow well and even obtained the highest net increment in dry weight (Table 2). As shown in Table 3, a large amount of NO₃-N was also present in the tissues of the NH₄NO₃-fed plants grown under the high nutrient level but not in the plants. These results support the conclusion that co-provision of can not only alleviate toxicity, but may also induce a synergistic growth response that can surpass the maximal growth rates produced by either N-source alone (Britto and Kronzucker, 2002). As long as there was , did not inhibit plant growth (Walch-Liu et al., 2000). One of the important roles of is as a signal to stimulate (or optimize) a multitude of biochemical responses (Tischner, 2000; Britto and Kronzucker, 2002).

Transpiration and water use efficiency

NO₃-N and NH₄NO₃ had similar effects on plant transpiration, while caused a marked reduction in transpiration (Fig. 1), although all of the treated plants were watered daily and showed no signs of water stress. The present results do not agree with the observation that use of increased leaf transpiration in alfalfa (Khan et al., 1994), tomato (Lugert et al., 2001), maize and wheat (Lewis et al., 1989), and increased stomatal conductance of white clover (Høgh-Jensen and Schjoerring, 1997). Although plants have smaller leaf area (Khan et al., 1994; Raab and Terry, 1994) and lower leaf area ratio (Magalhaes and Wilcox, 1984), these could not be the reasons to explain the reduced transpiration in the present study. At the second harvest, the shoots increased in dry weight by 3·15 g, 1·64 times more than that at the first harvest, and daily transpiration increased by only 33 %. The shoot increased in dry weight by 3·21 g, 1·67 times more than that at the first harvest, while daily transpiration was reduced by 13 % compared with the first harvest (Fig. 1). Furthermore, results of carbon isotope discrimination for different leaves of the NO₃- and plants indicated a reduced stomatal conductance in the plants, especially in the top and new leaves, and upper leaves, meaning that the stomata in the leaves were at least partly closed. Also in the NO₃-fed plants, a smaller stomatal conductance at the low nutrient level than that at the high nutrient level was observed (Fig. 2). The results suggest that the reduced leaf transpiration in the was not due to the reduced leaf area, but due to reduced stomatal conductance (Guo et al., 2002). There is some evidence that ABA may be responsible for the reduced stomatal conductance in leaves of the plants, since elevated biosynthesis of ABA and an obvious increase in the xylem and phloem flow of ABA in plants have been observed in comparison with NO₃-grown plants in castor beans (Peuke et al., 1994b; Jeschke and Hartung, 2000). On the other hand, Guo et al. (2003) provided evidence for a role of in stomatal opening, and demonstrated by using chl1 mutants of A. thaliana that the reduced stomatal aperture is not the result of effects on ABA responses. There is also an early report that supplies evidence that KCl supports stomatal opening as well as KNO₃ in Vicia epidermal strips (Humble and Hsiao, 1969).

N-sources can influence plant WUE and δ¹³C value. A common finding is that a lower WUE is correlated with a more negative δ¹³C value (Farquhar et al., 1989; Raven and Farquhar, 1990; Raven et al., 1992; Yin and Raven, 1997, 1998). Different nutrient levels may also influence WUE of plants. In winter oilseed rape (Brück et al., 2001) and sweet potato plants (Kelm et al., 2001), a low N level led to lower WUE, since N stress led to increased transpiration per unit leaf area and decreased WUE (Kelm et al., 2001). In the present study, however, opposite results were obtained. The WUE of the low-NO₃- and low-NH₄NO₃-fed plants was higher than that of the high-NO₃- and high-NH₄NO₃-fed plants (Table 5). By comparison of the transpiration rate on a leaf dry weight basis at the second harvest, it was found that the transpiration rate for , low-NO₃- and low-NH₄NO₃-fed plants was 18·6, and 17·8 g d⁻¹ g⁻¹ leaf dry weight, respectively, and for -, high-NO₃- and high-NH₄NO₃-fed plants it was 10·6, 25·0 and 23·9 g d⁻¹ g⁻¹ leaf dry weight, respectively (calculated based on the results in Table 2 and Fig. 1), showing a decreased transpiration rate for the low-NO₃- and low-NH₄NO₃-fed plants compared with that for the high-NO₃- and high-NH₄NO₃-fed plants. The WUE of the NO₃- or NH₄NO₃-fed plants was almost same under both nutrient levels (Table 5), which was in agreement with previous results for wheat (Yin and Raven, 1998) and peanut (Hubick, 1990). The WUE of the plants was the lowest and that of the plants the highest among the treatments. Apart from the negative influence on leaf stomatal conductance discussed above, the effect of on root water uptake rate relative to assimilation of carbon dioxide might also contribute to the difference in WUE under both nutrient levels. At the second harvest, the water uptake rate on a root dry weight basis for -, low-NO₃- and low-NH₄NO₃-fed plants was 76, 70 and 62 mL d⁻¹ g⁻¹ dry weight, respectively; i.e. the plants had the highest water uptake rate. For -, high-NO₃- and high-NH₄NO₃-fed plants the water uptake rate on a root dry weight basis was 45, 127 and 106 mL d⁻¹ g⁻¹ dry weight, respectively; i.e. the plants had the lowest water uptake rate, and the lowest leaf transpiration (Fig. 1).

Uptake, flow and partitioning of potassium

There was no difference in the amount of K⁺ taken up by NO₃- and NH₄NO₃-fed plants at the low nutrient level (2·5 mm potassium). By increasing the nutrient concentration (6 mm K; where the concentration of potassium supplied was 1·4 times more), potassium uptake increased by only 52 % and 75 % in the NO₃- and NH₄NO₃-fed plants, respectively (Table 4). This is in agreement with the proposed theory that cycling of potassium in plants can act as an important signal for feedback control of nutrient uptake (Drew et al., 1990; Engels and Marschner, 1992). However, when NO₃-N was replaced by , potassium uptake was substantially reduced, and it was and unchanged when the potassium supply was increased (high nutrient level). This is consistent with earlier results from castor bean (Van Beusichem et al., 1988) and wheat (Wang and Below, 1998), because of the strong inhibition of potassium uptake by (Marschner, 1995; Britto and Kronzucker, 2002).

not only inhibited potassium uptake, but also markedly influenced its flow and partitioning within plants. In the plants grown under both nutrient levels, the amount of potassium transported in the xylem was almost the same compared with the NO₃- and NH₄NO₃-fed plants. It was 457 % or 596 % of the potassium currently taken up at low or high nutrient level, respectively (Figs 3, 4), a situation that can be explained only by massive export from leaves and retranslocation of potassium in the phloem. Retranslocation of potassium from the shoot via the phloem and cycling through the roots also took place in the NO₃- and NH₄NO₃-fed plants (Figs 3, 4). It contributed to the potassium transported in the xylem by 52 %, 59 % and 80 % in the low-NO₃-, low-NH₄NO₃- and plants, respectively, and 41 %, 50 % and 84 % in the high-NO₃-, high-NH₄NO₃- and plants, respectively. Thus, retranslocation of potassium in the phloem of plants exceeded that of the NH₄NO₃- and NO₃-fed plants.

The results from the present study indicate that total replacement of NO₃-N by caused a substantial decrease in dry weight gain, even when plants were grown under nutrient deficiency. This suggests that the inhibitory effect of on plant growth might not be directly caused by ammonium toxicity. In addition, as the sole N-source caused a reduction in transpiration rate, changes in plant WUE (depending on the concentration of nutrient solution) and a decrease in potassium uptake. This leads to a massive export of K⁺ from leaves and recycling in the phloem. The results highlight the high degree of reutilization by retranslocation via the phloem, and provide further evidence that the partitioning and retranslocation of potassium in plants can be changed, depending on environmental conditions (Jeschke et al., 1987; Jeschke and Pate, 1991; Hibberd et al., 1999; Jiang et al., 2001).