Aluminium tolerance in rice is antagonistic with nitrate preference and synergistic with ammonium preference

Abstract

Background and Aims

Acidic soils are dominated chemically by more ammonium and more available, so more potentially toxic, aluminium compared with neutral to calcareous soils, which are characterized by more nitrate and less available, so less toxic, aluminium. However, it is not known whether aluminium tolerance and nitrogen source preference are linked in plants.

Methods

This question was investigated by comparing the responses of 30 rice (Oryza sativa) varieties (15 subsp. japonica cultivars and 15 subsp. indica cultivars) to aluminium, various ammonium/nitrate ratios and their combinations under acidic solution conditions.

Key Results

indica rice plants were generally found to be aluminium-sensitive and nitrate-preferring, while japonica cultivars were aluminium-tolerant and relatively ammonium-preferring. Aluminium tolerance of different rice varieties was significantly negatively correlated with their nitrate preference. Furthermore, aluminium enhanced ammonium-fed rice growth but inhibited nitrate-fed rice growth.

Conclusions

The results suggest that aluminium tolerance in rice is antagonistic with nitrate preference and synergistic with ammonium preference under acidic solution conditions. A schematic diagram summarizing the interactions of aluminium and nitrogen in soil–plant ecosystems is presented and provides a new basis for the integrated management of acidic soils.

Aluminium, ammonium, correlation, Indica, Japonica, nitrate, rice

INTRODUCTION

Aluminium (Al) is the most abundant metallic element in the Earth's crust (Yaroshevsky, 2006). Currently, there is no evidence that Al is essential for plant growth, although it is beneficial for some plant species (Pilon-Smits et al., 2009). Most Al in soil is present as harmless mineral forms. However, as soils become acidic, potentially toxic ionic forms of Al dissolve into the soil solution, inhibiting root growth and function. Consequently, Al toxicity may become the primary factor limiting crop production in acidic soils, which account for 50 % of the world's potentially arable lands (Kochian et al., 2005), whereas this toxicity is often lacking in neutral to calcareous soils.

Nitrogen (N) is the most abundant mineral nutrient element in plants, and plays an essential role in plant growth and development. Despite its relatively high concentrations and important functions in plants, N is present in very low concentrations in soil (Yaroshevsky, 2006). Ammonium (NH₄⁺) and nitrate (NO₃⁻) are the two main inorganic N sources available for plant uptake. In the field, inorganic N occurs predominantly as NH₄⁺ in soils of pH 4·0–6·0 and as NO₃⁻ in soils of pH 6·0–8·0 (McGrath and Rorison, 1982). Therefore, a supply of NH₄⁺-N becomes a critical factor for the survival of plants in acidic soils, whereas NO₃⁻-N does so in neutral to calcareous soils.

Thus, soil Al and N are two major pH-linked factors in plant nutrition. Acidic soils may be dominated chemically not only by Al but also by NH₄⁺, and neutral to calcareous soils, although lacking potentially toxic concentrations of Al, have higher concentrations of NO₃⁻ (Rorison, 1985). Soil characteristics can greatly influence the distribution of plant species or genotypes. Different plants vary greatly in their ability to absorb and utilize NH₄⁺- and NO₃⁻-N (Haynes and Goh, 1978) and in their tolerance to excess Al (Foy, 1988; Ma, 2005). Rice (Oryza sativa) is one of the world's most important crops, supplying food for nearly half its population, especially in Asia. About 13 % of global rice production occurs in acidic soils (von Uexküll and Mutert, 1995). On the whole, rice plants are the most Al-tolerant among small cereal crop species (Foy, 1988; Famoso et al., 2010). At the same time, because NH₄⁺ is the predominant N species in anaerobic agricultural soils, in particular in paddy fields, it has been generally accepted that NH₄⁺ is the most important source of N for rice plants, despite some reports that partial addition of NO₃⁻ to growth media can improve rice growth (Fan et al., 2005; Duan et al., 2007).

There are extensive reports on the responses of plants to N sources and Al toxicity, but N utilization and Al toxicity are often studied separately. Given the parallel association of Al and inorganic N forms with soil pH, plants may have evolved mechanisms that link Al stress and NH₄⁺-N nutrition as a result of ecological adaptation and natural selection. Thus, we hypothesized that: (1) Al-tolerant rice plants prefer NH₄⁺-N while Al-sensitive rice plants prefer NO₃⁻-N, and Al tolerance and N source preference may be linked in rice; and (2) Al can synergistically interact with NH₄⁺ but antagonistically with NO₃⁻ in their effects on rice growth. The present study was conducted in an attempt to test these hypotheses using various rice varieties supplied with various N and Al sources. The results obtained demonstrate coordination of inorganic N and Al in rice grown under acidic conditions.

MATERIALS AND METHODS

Plant materials and growth conditions

Fifteen genetically diverse varieties each of two rice subspecies (Oryza sativa subsp. japonica and O. sativa subsp. indica) were used (Table 1). Plants were grown in a controlled-environment growth chamber with day/night temperatures of 28 ± 1/20 ± 1 °C, a day length of 14 h, light intensity of 50 klux, and relative humidity of 65 ± 5 %. Seeds were surface-sterilized with 10 % H₂O₂ for 10 min, washed with deionized water several times, soaked for 24 h in deionized water in the dark, and then germinated for 24 h between filter papers soaked with deionized water. Subsequently, seedlings were transferred to a framed net floating on 0·5 mm CaCl₂ (pH 4·5) and grown for 2 d before use.

Table 1.

Rice varieties used: Nos. 1–15 are japonica, and Nos. 16–30 are indica

No.	Name	No.	Name	No.	Name
1	‘Nanjing55108’	11	‘Zhendao2’	21	‘Yangfuxian2’
2	‘Nanjing43’	12	‘Wuyujing14’	22	‘IIyou838’
3	‘Nipponbare’	13	‘Zaofeng9’	23	‘Zhenhui084’
4	‘Nanjing41’	14	‘86you8’	24	‘Yueyou277’
5	‘Koshihikari’	15	‘Wuyujing7’	25	‘Youmi7’
6	‘Suhuxiangjing’	16	‘Yangdao6’	26	‘Minghui70’
7	‘2723’	17	‘Liangyoupei9’	27	‘IR26’
8	‘Nanjing42’	18	‘Jin04-12’	28	‘Shengtai1’
9	‘Louming’	19	‘Zhongxian98’	29	‘Nanjing16’
10	‘Yandao8’	20	‘Teyou559’	30	‘Kasalath’

The Al³⁺, NH₄⁺ and NO₃⁻ were applied as AlCl₃.6H₂O, NH₄Cl and NaNO₃, respectively. At the outset of all experiments, the initial pH of the solutions was adjusted to 4·5 by the addition of 0·1 m HCl. The culture solutions were renewed every day in order to reduce the effects of pH changes on rice growth due to the uptake of NH₄⁺ and NO₃⁻.

Root elongation measurement

The roots of 2-d-old seedlings were exposed to 0·5 mm CaCl₂ (pH 4·5) with or without 50 µm Al for 24 h. The root length of each seedling was measured with a ruler before and after the treatments. Thirty rice varieties (Table 1) were used for root elongation measurement, and ten roots of each variety were measured. Relative root elongation was calculated as: (root elongation with Al treatment)/(root elongation without Al) × 100. Because different rice cultivars vary greatly in the absolute root elongation under no Al conditions, the relative root elongation was used here to compare the difference in Al tolerance of different rice cultivars.

N source preference evaluation

The 2-d-old seedlings were first grown in an N-free modified Kimura B and Arnon solution as described previously (Zhao et al., 2009). After 4 d of growth, uniform seedlings were selected and transplanted to a plastic pot containing 20 L of nutrient solution. The N sources were supplied as NH₄⁺ and NO₃⁻ at various NH₄⁺/NO₃⁻ ratios (100 : 0, 75 : 25, 50 : 50, 25 : 75 and 0 : 100) at an N concentration of 2 mm. After 24 d, roots and shoots were harvested separately, dried in a forced-air oven at 80 °C to constant weight, and weighed. Thirty rice varieties (Table 1) were used for N source preference evaluation, and there were two experimental replicates for each variety. Relative root dry weight was then calculated as follows: (root dry weight of rice seedlings at a specific NH₄⁺/NO₃⁻ ratio)/(root dry weight of rice seedlings at the NH₄⁺/NO₃⁻ ratio of 100 : 0) × 100; or (root dry weight of rice seedlings at a specific NH₄⁺/NO₃⁻ ratio)/(root dry weight of rice seedlings at the NH₄⁺/NO₃⁻ ratio of 0 : 100) × 100. Relative shoot dry weight was calculated in the same way, using shoots instead of roots. Because different rice cultivars vary greatly in the absolute biomass dry weight at the NH₄⁺/NO₃⁻ ratio of 100 : 0 or 0 : 100, the relative dry weight was used here to compare the difference in N source preference of different rice cultivars.

The interaction between Al and N

The 2-d-old seedlings were first grown in an N-free modified Kimura B and Arnon solution as described previously (Zhao et al., 2009). After 4 d of growth, uniform seedlings were then selected and exposed to nutrient solutions with or without 50 µm Al at various NH₄⁺/NO₃⁻ ratios (100 : 0, 75 : 25, 50 : 50, 25 : 75 and 0 : 100) at an N level of 2 mm for 24 d. During harvest, the roots of seedlings were washed with deionized water. Their roots and shoots were then harvested separately and dried in a forced-air oven at 80 °C to constant weight, weighed and ground for analysis of total N and Al concentrations in the plant tissues. In this experiment, ten rice varieties (Nos. 1, 2, 3, 5, 15, 16, 27, 28, 29, 30) (Table 1) were randomly chosen to investigate the interaction between Al and N, and there were three experimental replicates for each variety. Relative root dry weight increment was calculated as: (root dry weight of rice seedlings with Al treatment – root dry weight of seedlings without Al treatment)/(root dry weight of seedlings without Al treatment) × 100. Relative shoot dry weight increment was calculated using shoots instead of roots. Relative biomass dry weight increment was used here to compare the effects of Al on rice growth at different NH₄⁺/NO₃⁻ ratios.

During the rice culture period, NH₄⁺ concentrations in the solution with NO₃⁻ treatment and NO₃⁻ concentrations in the solution with NH₄⁺ treatment were negligible, so it can be assumed that N was absorbed by rice plants in the forms supplied.

Analysis of total N and Al concentrations in plant tissue

A total of 100–500 mg of the ground samples were digested with 5 mL of concentrated HNO₃ to analyse root and shoot Al concentrations as described by Shen (2008). Al contents in the digested solutions were determined after appropriate dilution by inductively coupled plasma-atomic emission spectroscopy (ICP-AES; IRIS-Advantage, Thermo Elemental, Franklin, MA, USA). For the analysis of N concentrations in roots and shoots, 100–200 mg of the ground samples was digested with 5-mL aliquots of a mixture of concentrated H₂SO₄ and 30 % H₂O₂, added to the digestion solution occasionally until the solution became clear. The total Kjeldahl N concentrations in the digested solutions were determined after appropriate dilution via a continuous flow analyser (AutoAnalyzer 3, Bran & Luebbe, Norderstedt, Germany).

Statistical analysis

The pots used for rice growth were randomly placed in the growth chamber after rice seedlings were transplanted, and the positions of all pots were changed every day during the subsequent growth period in order to reduce the influence of possible light non-homogeneity. A mean comparison was made according to independent-samples t-test (two-range comparison) or Tukey's test (multi-range comparison) using the Statistical Analysis System (SPSS 18·0).

RESULTS

Comparison of Al tolerance between the japonica and indica subspecies

The most direct and obvious effect of Al toxicity on plant growth is the inhibition of root elongation after only several minutes or hours by micromolar concentrations of Al. Therefore, relative root elongation is considered to be the parameter most directly related to Al tolerance in plants (You et al., 2005). Here, relative root elongation was also used to compare Al tolerance between the japonica and indica subspecies. The relative root elongation of the japonica varietal group was significantly higher than that of the indica group (Fig. 1), indicating that the former was more tolerant to Al than the latter.

Fig. 1.

Comparison of Al tolerance in the japonica and indica varietal groups using relative root elongation of rice seedlings. Two-day-old rice seedlings were exposed to 0·5 mm CaCl₂ (pH 4·5) with or without 50 µm Al in the absence of N supply for 24 h. Data are means ± s.d. (n = 15) for 15 japonica varieties or 15 indica varieties as shown in Table 1. The method used to calculate relative root elongation is described in the Materials and Methods. *Indicates a statistically significant difference between japonica and indica (P < 0·05 by independent-samples t-test).

Comparison of N source preference between the japonica and indica subspecies

Moderate NH₄⁺- and NO₃⁻-N concentrations generally have small short-term influences on the root elongation of plants, so changes in the root elongation of rice seedlings are not suitable for assessing N source preference in rice. It is somewhat difficult to compare the N source preference among different plant species and cultivars within the same species because it is not easy to find a uniform evaluating index as with the relative root elongation in evaluating Al tolerance.

Relative root dry weight and relative shoot dry weight were employed to compare the N source preference between japonica and indica subspecies by assigning the dry weight at the NH₄⁺/NO₃⁻ ratio of 100 : 0 a value of 100. The results showed that the relative root dry weight and relative shoot dry weight of the indica varietal group were significantly higher than those of the japonica varietal group at NH₄⁺/NO₃⁻ ratios of 75 : 25, 50 : 50, 25 : 75 and 0 : 100 (Fig. 2A, B). When the dry weight of seedlings at the NH₄⁺/NO₃⁻ ratio of 0 : 100 was assigned a value of 100, the relative root dry weight and relative shoot dry weight of the japonica varietal group were significantly higher than those of the indica varietal group at the NH₄⁺/NO₃⁻ ratio of only 100 : 0, but there was no significant difference in the relative root dry weight and relative shoot dry weight between japonica varietal group and indica varietal group at the NH₄⁺/NO₃⁻ ratios of 75 : 25, 50 : 50 and 25 : 75 (Fig. 2C, D). These results demonstrated that the indica group performed better in the growth medium containing NO₃⁻ than did the japonica group, whereas the latter grew better in the medium supplied solely with NH₄⁺-N than did the former.

Fig. 2.

Comparison of N source preference in the japonica and indica varietal groups using relative root dry weight (A, C) and relative shoot dry weight (B, D). Six-day-old rice seedlings were grown in nutrient solutions with various NH₄⁺/NO₃⁻ ratios (as indicated) at an N concentration of 2 mm in the absence of Al treatment for 24 d. (A, B) Dry weight at the NH₄⁺/NO₃⁻ ratio of 100 : 0 was assigned a value of 100, whereas in (C, D) dry weight at the NH₄⁺/NO₃⁻ ratio of 0 : 100 was assigned a value of 100. Data are means ± s.d. (n = 15) for 15 japonica varieties and 15 indica varieties, as shown in Table 1. The method to calculate relative root dry weight and relative shoot dry weight is described in the Materials and Methods. Different letters above the column indicate statistically significant differences among different columns (P < 0·05 by Tukey's test).

The different NH₄⁺/NO₃⁻ ratios markedly affected the growth of the indica group, and the relative root dry weight and relative shoot dry weight of the indica varietal group gradually increased from 100 : 0 to 25 : 75 and then significantly decreased at 0 : 100 (Fig. 2), suggesting the indica group preferred certain ratios of NH₄⁺ and NO₃⁻ to single NO₃⁻ or NH₄⁺ although it showed better growth with single NO₃⁻ than single NH₄⁺. However, different NH₄⁺/NO₃⁻ ratios had much less influence on the relative root dry weight and relative shoot dry weight of the japonica group (Fig. 2). Therefore, growth of the indica group was affected more by differences in NH₄⁺/NO₃⁻ ratios than the japonica group.

Correlation between Al tolerance and N source preference

Simple linear correlation analysis between Al tolerance and N source preference showed that, when the dry weight at the NH₄⁺/NO₃⁻ ratio of 100 : 0 was assigned a value of 100, relative root elongation of rice plants was significantly negatively correlated with both relative root dry weight and relative shoot dry weight of rice plants grown at NH₄⁺/NO₃⁻ ratios of 75 : 25, 50 : 50, 25 : 75 and 0 : 100 (Fig. 3A–H), indicating that Al tolerance is significantly negatively correlated with a preference of rice seedlings for NO₃⁻-N. Nevertheless, when the dry weight of seedlings supplied with N at the NH₄⁺/NO₃⁻ ratio of 0 : 100 was assigned a value of 100, relative root elongation of rice plants was significantly positively correlated with both relative root dry weight and relative shoot dry weight of rice plants grown at the NH₄⁺/NO₃⁻ ratio of 100 : 0 (Fig. 3I, M) but not at ratios of 75 : 25, 50 : 50 and 25 : 75 (Fig. 3J–L, N–P), indicating that Al tolerance was positively correlated with a preference of rice seedlings for NH₄⁺-N only at the NH₄⁺/NO₃⁻ ratio of 100 : 0.

Fig. 3.

Correlation of Al tolerance and N source preference across 30 genetically diverse rice varieties. The concentration of Al was 50 µm (Fig. 1). The concentration of N was 2 mm (Fig. 2). Correlation analysis between relative root elongation and relative root dry weight (A–D, I–L) or relative shoot dry weigh (E–H, M–P) was done at NH₄⁺/NO₃⁻ ratios of 100 : 0, 75 : 25, 50 : 50, 25 : 75 and 0 : 100, as indicated on the graphs, using 30 rice varieties as shown in Table 1. (A–H) Dry weight at the NH₄⁺/NO₃⁻ ratio of 100 : 0 was assigned a value of 100; (I–P) dry weight at the NH₄⁺/NO₃⁻ ratio of 0 : 100 was assigned a value of 100. Each data point represents the value for each rice variety of 15 japonica varieties and 15 indica varieties (indicated in the key in A), as shown in Table 1. The methods to calculate relative root elongation, relative root dry weight and relative shoot dry weight are described in the Materials and Methods. *Significant correlations between Al tolerance and N source preference among the different rice varieties at the 1 % level (n = 30).

Interaction between Al and N source supply

After comparing Al tolerance and NH₄⁺ or NO₃⁻ preference among the 30 rice varieties, ten rice varieties, including five japonica (Nos. 1, 2, 3, 5 and 15 in Table 1) and five indica (Nos. 16, 27, 28, 29 and 30 in Table 1) varieties, were randomly chosen for further study of the interaction between Al and inorganic N.

The relative dry weight increment of biomass was used to evaluate the effects of Al on rice growth at different NH₄⁺/NO₃⁻ ratios. When supplied with only NH₄⁺-N (NH₄⁺/NO₃⁻ ratio of 100 : 0), the values of relative root dry weight increment and relative shoot dry weight increment of both japonica and indica varietal group were positive, but they were negative when only nitrate was supplied (NH₄⁺/NO₃⁻ ratio of 0 : 100) (Fig. 4A, B), indicating that rice growth was severely inhibited by Al when supplied with only NO₃⁻-N but improved by Al when supplied with only NH₄⁺-N. Aluminium addition did not generally affect rice growth at other NH₄⁺/NO₃⁻ ratios (75 : 25, 50 : 50 and 25 : 75; data not shown).

Fig. 4.

The effect of Al on the growth of japonica and indica rice plants grown with NH₄⁺ or NO₃⁻ alone. Six-day-old rice seedlings were grown in nutrient solutions with or without 50 µm Al at NH₄⁺/NO₃⁻ ratios of 100 : 0 or 0 : 100 at an N level of 2 mm for 24 d. The calculation method of relative root dry weight increment and relative shoot dry weight increment is described in the Materials and Methods. (A) Relative root dry weight increment; (B) relative shoot dry weight increment; (C) root Al concentrations; (D) shoot Al concentrations. Data are means ± s.d. (n = 5) for five japonica varieties or five indica varieties as described in the Materials and Methods. The data of Al concentrations of roots and shoots for the absence of Al are not shown due to very small values. Different letters above the column indicate statistically significant differences among different columns (P < 0·05 by Tukey's test).

Regardless of japonica or indica group, nitrate-fed rice seedlings accumulated about 30 times more Al in roots than plants supplied with NH₄⁺-N (Fig. 4C). The shoot Al concentration of the japonica varietal group was significantly higher at NH₄⁺/NO₃⁻ ratio of 0 : 100 than 100 : 0 while that of indica was not (Fig. 4D). The Al concentrations of roots at other NH₄⁺/NO₃⁻ ratios (75 : 25, 50 : 50 and 25 : 75) increased as NO₃⁻-N ratios increased in the growth medium (data not shown). These results indicated that Al accumulation in roots was greatly increased by NO₃⁻ in comparison with NH₄⁺.

No significant differences in the N concentrations of roots and shoots were observed in the japonica and indica groups between –Al and +Al treatments, or with different NH₄⁺/NO₃⁻ ratios (data not shown). Nevertheless, because Al increased the dry weights of NH₄⁺-fed rice seedlings but decreased those of NO₃⁻-fed rice seedlings (Fig. 4A, B), Al increased N uptake (N concentration × dry weight) of NH₄⁺-fed rice seedlings while it decreased it in NO₃⁻-fed seedlings (calculated results not shown).

DISCUSSION

Rice is the most Al-tolerant among small cereal crop species (Foy, 1988; Famoso et al., 2010). In this study, we have further demonstrated that japonica is more tolerant to Al toxicity than the indica group (Fig. 1), consistent with previous reports (Ma et al., 2002; Watanabe and Okada, 2005; Chen and Shen, 2008; Yang et al., 2008; Famoso et al., 2010). Famoso et al. (2010) hypothesized that, over the course of evolution, Oryza experienced a dramatic shift in its position within the range of plant responses to Al, which led to dramatic genetic change and enhanced Al tolerance. The possibility of Al toxicity in soils may be an important factor in the distribution of plant species, subspecies and varieties. Therefore, we can infer that the japonica group may have grown originally in an environment with more, while indica may have grown in an environment with less, potential for Al toxicity.

Although NH₄⁺ is generally accepted to be the preferred N source of rice, different conclusions have often been obtained depending on experimental conditions and rice varieties. The indica group utilizes NO₃⁻ as an N source more efficiently than japonica (Ta et al., 1981; Fan et al., 2005), and this superiority can be influenced by various environmental and medium conditions (Ta and Ohira, 1981, 1982a, b). The present study confirms the superiority of the indica group in utilizing NO₃⁻-N compared with japonica (Fig. 2A, B). However, in previous reports (Ta et al., 1981; Ta and Ohira, 1981, 1982a, b), the japonica and indica groups responded similarly to NH₄⁺-N supply, which is inconsistent with the present study, where the japonica group grew better under a sole NH₄⁺-N supply than did the indica group (Fig. 2C, D). This discrepancy may be attributed to the pH medium used. Ta et al. (1981) used a medium of pH 4·5, 4·5 and 6·0, respectively, in the NO₃⁻, NO₃⁻ + NH₄⁺ and NH₄⁺ treatments. In contrast, the initial pH of the medium we used was 4·5 in all N treatments. Improved rice growth at higher pH may reduce or mask differences in NH₄⁺-N utilization between japonica and indica varietal groups previously reported. Results in this study demonstrate that indica group plants grew better in the medium containing only NO₃⁻ or NH₄⁺–NO₃⁻ mixtures than did the japonica group plants while the latter did better only in conditions of NH₄⁺-N supply. Thus, as with Al, it is inferred that the japonica varietal group originally grew in environments with more NH₄⁺-N while the indica group occupied environments with more NO₃⁻-N.

Plant species adapted to acidic soils have a preference for NH₄⁺-N while plants with a preference for neutral to calcareous soils preferentially utilize NO₃⁻-N (Rorison, 1985; Falkengren-Grerup, 1995; Marschner, 1995). Accordingly, the preference for NH₄⁺-N or NO₃⁻-N in plants should be closely related to soil pH. Thus, acid-sensitive plants should prefer a NO₃⁻-N source, but acid-tolerant plants should prefer an NH₄⁺-N source. Because toxic Al levels in soils are mainly affected by pH and a decrease in pH lower than 5·0 can rapidly increase Al to toxic concentrations, it is reasonable to guess that the preference of plants for NH₄⁺-N and NO₃⁻-N should have some relationship with Al tolerance. This suggestion is supported by the data shown here, in which Al tolerance in the rice varieties was significantly negatively with a preference for NO₃⁻-N (Fig. 3). This information can be used to screen for rice varieties that are both Al-tolerant and NH₄⁺-efficient. If such rice varieties can be grown in acidic soils, Al tolerance and N utilization might be improved simultaneously. This may be useful for enabling integrated management of acid soils and increasing plant and grain production in the future.

The indica and japonica subspecies, the two major varieties of cultivated rice, have distinct eco-geographical distribution ranges (Zhang et al., 1992). Rice subspecies differ in many trait performances with respect to biotic and abiotic stress (Famoso et al., 2010); indica rice is predominant in tropical and subtropical regions that comprise the major part of the global rice-growing areas, while japonica rice is more adapted to temperate regions (Zhang et al., 1992). However, acidic soils are widely distributed in humid zones of tropical, subtropical and temperate regions (von Uexküll and Mutert, 1995). This indicates that japonica and indica are not distributed entirely according to soil pH. Therefore, there must be other factors other than N and Al driving the differentiation of japonica and indica rice.

The data presented here demonstrate that Al is beneficial to rice growth in a medium containing NH₄⁺-N, but is harmful in an NO₃⁻-N medium (Fig. 4). This suggests a synergistic effect between NH₄⁺ and Al, but an antagonistic effect between NO₃⁻ and Al. There is extensive literature on the alleviating effect of NH₄⁺ on Al toxicity in other plant species (e.g. McCain and Davies, 1983; Rorison, 1985; Klotz and Horst, 1988; Cumming, 1990; Cumming and Weinstein, 1990; Grauer and Horst, 1990; Schier and McQuattie, 1999). We have previously demonstrated that NH₄⁺-N can alleviate Al toxicity in rice and Lespedeza compared with NO₃⁻-N. The mechanism is that NH₄⁺ ions decrease the cation-binding sites of the cell wall through direct NH₄⁺ and indirect proton competition with Al ions (Zhao et al., 2009; Chen et al., 2010).

Interestingly, in this study, Al increased the dry weights of rice supplied with NH₄⁺-N, which is consistent with previous reports on grasses (Rorison, 1985) and tropical trees (Watanabe et al., 1998). The synergism between Al and NH₄⁺ can be understood in two ways: one is that Al increases the NH₄⁺-N utilization of rice, and the other is that Al decreases possible NH₄⁺ toxicity in the rice, as higher concentrations of NH₄⁺-N are often toxic to plants (Britto and Kronzucker, 2002). The mechanisms responsible for the improved rice growth in the presence of Al and an NH₄⁺-N supply cannot be elucidated from the present data. One reason for this may be that Al ions reduced the toxicity of protons secreted from rice roots after NH₄⁺-N was taken up, as the Al³⁺ and H⁺ ions are mutually antagonistic (Kinraide, 2003). In the present study, the pH of the NH₄⁺-N and NO₃⁻-N media changed from an initial 4·5 to final values of 3·5–3·8 and 4·5–5·5, respectively, after 24 h of culture, although the nutrient solutions were renewed every day. Another possible reason is that Al can increase the activity of NH₄⁺-N assimilatory enzymes. Some reports indicate that Al activated chloroplastic glutamine synthetase in wheat (Pécsváradi et al., 2009) and enhanced the expression of the gene encoding glutamine synthetase in rice (Zhang et al., 2007). The activation of glutamine synthetase can transform NH₄⁺-N into glutamine, so NH₄⁺ toxicity was alleviated in the presence of Al compared with in the absence of Al. Thus, rice growth was increased under the combination of Al and NH₄⁺-N relative to the only NH₄⁺-N supply.

It is generally accepted that wheat and barley are both NO₃⁻-N-preferring and Al-sensitive, while tea trees are NH₄⁺-N- and Al-preferring. On the basis of this and our results, we summarized the interactive behaviour of Al and N in soil–plant ecosystems by a schematic diagram (Fig. 5). From this diagram, we observe that neutral and calcareous soils are chemically dominated by lower Al³⁺ and higher NO₃⁻ concentrations, while acidic soils are characterized by more Al³⁺ and more NH₄⁺. Correspondingly, plants originally growing in neutral and calcareous soils might be both Al-sensitive and NO₃⁻-N-preferring while plants originally growing in acidic soils might be both Al-tolerant and NH₄⁺-N-preferring. Moreover, NO₃⁻ decreases, while NH₄⁺ increases, toxic Al concentration in soils as NO₃⁻ increases while NH₄⁺ decreases soil pH due to their uptake and N transformation. However, within plants, Al and NO₃⁻ are antagonistic while Al and NH₄⁺ are synergistic. Thus, plants are able to make full use of various inherent mechanisms to resolve the different situations in soils. However, why the NO₃⁻-preferring and Al-sensitive indica rice can extensively distribute in acid soils remains unclear. Therefore, there are other important factors such as temperature and water to drive the geographical distribution of plants than Al and N.

Fig. 5.

Schematic diagram of the relationship between N and Al in soil–plant ecosystems with regard to soil pH. The top half above the line indicates N and Al behaviours within plants. The bottom half below the line indicates N and Al behaviours within soils. The left side represents neutral to calcareous soils; the right side represents acid soils. The circles indicate the characteristics of the plants and soils, whilst the rectangles indicate the N–Al interactions within them.

It should be emphasized that the schematic diagram shown in Fig. 5 is suitable for most plant species and soil types, but some particular plants and soils may be excluded from the predicted patterns. More experimental results are needed to test and develop this. It was mainly developed based the results of our research described here. To test it further, the relationships between Al tolerance and N source preference need to be investigated, both with different plant species and with different varieties of the same plant species. In addition, the effects of Al on the chemical processes of NH₄⁺ and NO₃⁻ in soils, especially nitrification, remain poorly known. This information is essential if we are to identify the role of Al in the global N cycle.

Finally, we conclude that Al and NH₄⁺ show synergistic behaviour while Al and NO₃⁻ are antagonistic in rice under acid conditions. This close N–Al relation in plants will provide integrated knowledge enabling increases in plant production by increasing the coordinated adaptation of plants to NH₄⁺ and Al coexisting in acid soils.

ACKNOWLEDGEMENTS

This work was supported by the National Natural Science Foundation of China (No. 41025005, 31000933), and National Natural Science Foundation of China (NSFC) – Japan Science and Technology Agency (JST) Cooperative Research Project (No. 30821140538). We thank Philip C. Brookes (Rothamsted Research) for his kind assistance in revising this paper, Yongchun Zhang (Jiangsu Academy of Agricultural Science) and Hongsheng Zhang (Nanjing Agricultural University) for their generous supply of rice seeds, and Ge Song (Institute of Soil Science, Chinese Academy of Sciences) for her assistance in determining N concentrations in rice.