Correspondence: T. Yoneyama Department of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, Bunkyo-ku, Tokyo 113–8657, Japan. Fax: + 81 3 5841 8032
Studies of uptake of ionic sources of N by two hydroponically grown rice (Oryza sativa L.) cultivars (paddy-field-adapted Koshihikari and dryland-adapted Kanto 168) showed that the magnitude of the nitrogen isotope fractionation (ɛ) for uptake of NH4+ depended on the concentrations of NH4+ and cultivar (averaging –6·1‰ for Koshihikari and –12·0‰ for Kanto 168 at concentrations from 40 to 200 mmol m−3 and, respectively, –13·4 and –28·9‰ for the two cultivars at concentrations from 0·5 to 4 mol m−3). In contrast, the ɛ for uptake of NO3− in similar experiments was almost insensitive to the N concentration, falling within a much narrower range (+3·2‰ to –0·9‰ for Koshihikari and –0·9‰ to –5·1‰ for Kanto 168 over NO3− concentrations from 0·04 to 2 mol m−3). From longer term experiments in which Norin 8 and its nitrate-reductase deficient mutant M819 were grown with 2 or 8 mol m−3 NO3− for 30 d, it was concluded that the small concentration-independent isotopic fractionation during absorption of this ion was not related to nitrate reductase activity.
Plants and microbes use inorganic nitrogen to synthesize organic nitrogen compounds. The major inorganic nitrogen forms absorbed by plants are NH4+-N and NO3−-N. In dryland crop fields, NH4+-N concentrations in the soil solutions range from 10 to 500 mmol m−3 and NO3−-N concentrations from 100 mmol m−3 to 10 mol m−3. In paddy (waterlogged) fields the concentrations of NH4+-N range from 100 mmol m−3 to 5 mol m−3 (Yoneyama T., unpublished). Previous studies indicated variations in natural nitrogen isotope abundance (δ15N) of organisms, caused by differences in the δ15N of their N source and metabolically induced changes of N isotope ratios (Handley & Raven 1992; Yoneyama 1995).
Studies of marine bacteria and plankton correlated δ15N values of these organisms with the concentrations of available N sources: NH4+, NO3− and CO(NH2)2 (Hoch, Fogel & Kirchman 1992; Pennock et al. 1996; Waser et al. 1998). The N concentrations in organisms' environments vary among microsites as well as on macro scales. If an organism exhausts a N source completely, the δ15N of that organism will be very close to that of the N source, but, if an organism absorbs only part of the source, nitrogen isotope fractionation may affect both the δ15N value of the organism and the δ15N value of the N remaining in the source without disturbing the overall isotopic mass balance.
Few studies of higher plants have considered the dependency of whole-plant N isotope fractionation during N uptake on N source concentrations. When presented at the level of mol m−3, the N isotopic fractionation associated with the uptake of NO3−-N ranged from 0 to 3·3‰ for the C4 grass, Pennisetum (Mariotti et al. 1982), ranged from 0 to −9·5‰ (Kohl & Shearer 1980) or was +2·1‰ (Bergersen, Peoples & Turner 1988) for the C3 legume, soybean, and was zero for C3 brassica, komatsuna (Yoneyama & Kaneko 1989). Measured against a common isotopic source of NO3−-N Handley et al. (1997) and Robinson et al. (2000) found significant genotypic variation and interaction of genotypic variation with abiotic stresses for shoot and whole-plant δ15N of cultivated and wild barleys.
NH4+-N uptake by rice caused an isotopic fractionation of –4·1 to –7·5‰ and –7·8 to –12·6‰ at external NH4+-N concentrations of 1·4 and 7·1 mol m−3, respectively (Yoneyama et al. 1991). In contrast, uptake of NH4+-N or NO3−-N by plants of a tomato cultivar (C3) supplied with 50 mmol m−3 did not result in whole-plant fractionation relative to source (Evans et al. 1996). Under NH4+-N nutrition, the tomato plants grew poorly and accumulated little biomass. On the basis of these results, Evans et al. (1996) concluded that no N isotope fractionation between plant and source had occurred under either NO3−-N or NH4+-N nutrition at N concentrations typically found in the field.
It has been hypothesized, but not conclusively demonstrated for NO3−-N, that previously observed fractionations (Mariotti et al. 1982) between plant δ15N and source δ15N depend on the ratio of demand by assimilatory enzyme (enzyme activity) to the concentration of the external N supply, and that a high external N concentration relative to enzymatic demand leads to isotopic fractionation between whole-plant and source. Mariotti et al. (1982) found a correlation between N isotope fractionation and plant age, which was also related with the development of NO3− reductase activity. This single report of this relationship in higher plants under NO3−-N nutrition may be confirmed independently by comparing the N isotope fractionation for NO3− uptake between normal plants and their NO3− reductase-deficient mutants.
In the present study we set out: (1) to determine the dependency of N isotope fractionation in whole plants on N concentrations that are realistic for field conditions, using contrasting cultivars of rice which are adapted to either substantial NH4+-N nutrition in paddy (waterlogged) or predominantly NO3−-N nutrition in dryland conditions and (2) to determine the effects of NO3− reductase activity on the N isotope fractionation using a normal cultivar versus its NO3− reductase deficient mutant.
MATERIALS AND METHODS
Seeds of paddy (waterlogged) rice cultivar Koshihikari and dryland cultivar Kanto 168 were germinated on moist filter paper at 25 °C in the dark. Five days after germination, they were transferred to a 0·25 mol m−3 CaSO4 solution. After a further 9 d, the seedlings were transferred to 3·5 L pots, six plants of a cultivar per pot, containing a hydroponic medium of the following composition: 0·25 mol m−3 Ca2+, 0·5 mol m−3 Mg2+, 0·25 mol m−3 PO43−, 0·5 mol m−3 K+, 1 mmol m−3 Cu2+, 1 mmol m−3 Zn2+, 1 mmol m−3 Mo6+, 10 mmol m−3 Mn2+, 30 mmol m−3 B3+, 100 mmol m−3 Fe (as EDTA) and from 1 to 2 mol m−3 NH4+ depending on the growth rate. Ten pots were prepared for each cultivar. The nutrient solution was changed three times per week, its pH was adjusted to 6·5 every day, and the solution gently bubbled with air throughout the entire experiment.
When the plants were 22 d old, the medium in the pots was changed for new medium containing either 2 mol m−3 NO3− or 2 mol m−3 NH4+, five replicate pots of each medium for both cultivars. These media were supplied at this concentration for a further 14 d for Kanto 168 and 35 d for Koshihikari. After this period, the plants were rinsed in fresh N-free solution for 1 h. Subsequently, the plants were fed for 30–120 min with medium (approx. 3 L) containing 100 mmol m−3 FeCl2 instead of Fe(EDTA), and either NO3− or NH4+, at sequentially increasing concentrations; 40, 50, 80, 100, 200, 500, 1000, 2000 and 4000 mmol m−3 (2000 mmol m−3 for NO3−). Medium pH for each treatment was initially adjusted to 6·5 ± 0·1. Between concentrations, plants were rinsed in N-free medium for 1 h. After 30 min of feeding for 40, 50 and 80 mmol m−3, 30 and 60 min for 100 and 200 mmol m−3, and 60 and 120 min for higher N concentrations, 500 mL of the medium was collected for analysis of its N-concentration and δ15N value. The amounts and δ15N values of N absorbed by the rice plants during 30–120 min feeding periods with 40–4000 mmol m−3 N were calculated by isotopic mass balance, using the measured amounts and δ15N values of N that remained in the medium. Every N treatment was conducted with five replicates. However, only the pots, in which more than 250 μg N remained in the pot after uptake or more than 10% of the added N was absorbed by the plants, were included in the results (Tables 2–5). The biomass of shoots and roots used for experiment 1 were measured after the final treatments.
Table 2. Absorption of ammonium and 15N/14N fractionation by Koshihikari plants at different initial NH4+ concentrations. Koshihikari is a dryland cultivar
Seeds of paddy rice cultivar, Norin 8 and its nitrate-reductase deficient mutant, M819 (Hasegawa et al. 1992) were germinated on filter paper. Five-day-old seedlings were transferred to a floating net in N-free 1 mol m−3 CaSO4 solution to culture for a further 10 d in a naturally lit glass house. Fifteen-day-old plants were transferred to pots with 3·5 L nutrient solution with the same composition as the initial hydroponics medium described in experiment 1, except for N and Fe. Nitrogen was supplied either as 2 or 8 mol m−3 KNO3. Iron was supplied as 100 mmol m−3 Fe-citrate, and to ameliorate the chlorosis, a 18 mol m−3 FeSO4·7H2O solution was sprayed on the leaves of 20-, 30-, and 40-day-old plants. The media were bubbled with air, and their pH was adjusted to 6·5 ± 0·1 every 3 d. Three replicated pots with two plants per pot were prepared, both for Norin 8 and M819. The nutrient solution was changed when plants were 30 (15 d of NO3−-N feeding) and 40 (25 d of NO3− -N feeding) days old. Fifteen (just before NO3−-N feeding) and 45-day-old plants (after 30 d of NO3−-N feeding) were harvested and, after washing with distilled water, freeze-dried. After grinding the freeze-dried samples, plant NO3− was extracted with hot water and purified over a cation exchange resin as described by Yoneyama & Kaneko (1989). The NO3− thus extracted was trapped in a 50 mmol m−3 H2SO4 solution, after sequential addition of MgO, to alkalize the solution, and Devarda's alloy, to reduce NO3− to NH4+. In this experiment, accumulation of whole-plant N during the 30 d of NO3−-N feeding was calculated as the difference of N between 15- and 45-day-old plants.
Analysis of N contents and δ15N values
After adjusting the solution pH to 7, aliquots of the collected culture solutions from experiment 1 and of ammonia converted from nitrate in experiment 2, each containing approximately 250 μg N, were concentrated by evaporation on a hot plate to 0·5 mL. Of this concentrate, 0·1 mL was transferred to a tin capsule and freeze-dried. The N-content and δ15N value of the freeze-dried samples were measured by an on-line ANCA-SL mass-spectrometer (Europa Scientific, Crewe, UK) using single N mode. For NH4+ samples, O2 was present during combusition in the analyses, but for NO3− samples, O2 was excluded from the furnace because NO3− is already the highest oxidation state of N. Approximately 5 mg of each ground plant sample from experiment 2 was weighed into a tin capsule and its N content and δ15N value were analyzed as described above.
The natural abundance of nitrogen isotopes (δ15N) in parts per thousand (‰) was calculated by Eqn 1,
where R = 15N/14N. Atmospheric N2 was used as the standard. Analytical error (SD) was within ±0·2‰.
The isotope fractionation factor (ɛ) for whole plants relative to source was calculated, assuming Rayleigh distillation, as
( (2) )
in which Rt = [δ 15N SIN/103 + 1] at time t, R0 is R at the start of the experiment and is the fraction of initially supplied inorganic N (SIN) that remained in the medium.
For the long-term experiment with Norin 8 and M819, the approximate isotopic fractionation accompanying NO3− uptake and NO3− reduction were calculated as
( (3) )
Where δ15Ns is the δ15N value of the N source and δ15Np is the δ15N value of the plant or of reduced N.
Shoots of Koshihikari, adapted to paddy conditions, grew larger with NH4+ than with NO3−; those of Kanto 168, adapted to dryland conditions, were slightly larger when grown on NO3− medium. Root growth of both varieties was less in NH4+ than in NO3− (Table 1). The calculated ɛ for NH4+ and NO3− uptake is given in Tables 2 and 4 for Koshihikari and in Tables 3 and 5 for Kanto 168. At low concentrations (40–200 mmol m−3), the value of ɛ ranged between −2 and −8·1‰ for NH4+ absorption by Koshihikari and the difference in δ15N between supplied NH4+ (−8·1‰) and absorbed N was 0·3 to 4·7‰. At high NH4+ concentrations (0·5–4 mol m−3), values of ɛ ranged between −9·4 and −18‰ for the same cultivar. For this concentration range the difference in δ15N between supplied NH4+ and absorbed N (−4·5 to −11·6‰) was larger than for the low concentrations. For NH4+ absorption by Kanto 168, the dryland cultivar, larger ɛ values (Table 3) were found than for Koshihikari, namely between −3·9 and −24·1‰ at low-to-medium N concentrations (up to 0·2 mol m−3) and between −15·3 and −43·2‰ for concentrations above 0·5 mol m−3. The relative depletion in 15N abundance of the absorbed N, when compared to supplied N was 1·2 to 13·2‰ at low-to-medium concentrations and 6·1 to 38·6‰ at high concentrations. Although there were some overlapping values, it seems obvious that in general for both cultivars ɛ values and observed differences in δ15N between absorbed and supplied N were more negative for media with high N concentrations than for media with low N concentrations. In contrast to NH4+ absorption (Tables 2 and 3), over a wide range of medium NO3− concentrations (40–2000 mmol m−3 NO3−) little variation in ɛ value was found for NO3− absorption by both cultivars (Tables 4 and 5). It is worth noting that Koshihikari showed some positive ɛ values at low and high medium NO3− concentrations which indicate net discrimination in favour of 15N rather than 14N during absorption.
Table 1. Fresh weights of rice plants used for short-term N uptake (expt 1)
39 ± 4
58 ± 9
24 ± 2
40 ± 4
Means of fresh weight per pot ± SE of five pots. Each pot contained six plants. Koshihikari was 59 d old and Kanto 168 was 38 d old.
152 ± 15
98 ± 18
46 ± 2
57 ± 3
In the long-term experiments, nutrient solutions were changed only twice (30 and 40-day-old) during NO3− feeding. This was done to increase the changes in δ15N values of plant N under large doses of medium N. However, only small changes in δ15N were observed for whole-plant N and none for the culture medium N (data not shown). Norin 8 plants had absorbed less N than the nitrate-reductase deficient mutant M819 at the end of the experiment, regardless of the initial NO3− concentration (2 or 8 mol m−3; Table 6). However, more of this absorbed N was recovered as NO3− in the mutant than in Norin 8, indicating that the rate of nitrate reduction was lower in M819 than in Norin 8. N isotope fractionation (Δ) during NO3− uptake was similar for both strains. Uptake from a 2 mol m−3 NO3− medium resulted in a positive fractionation value, in contrast to uptake in cultures with 8 mol m−3 NO3−, where the fractionation value was negative (Table 6). Nitrogen isotope fractionation values for nitrate reduction in the plants, calculated by δ15N of reduced plant N –δ15N of unassimilated NO3−-N (Ledgard, Woo & Bergersen 1985), were more negative for M819 than for Norin 8. The values for plants cultured with 2 mol m−3 NO3− were 4 to 5‰ less negative than for the plants cultured with 8 mol m−3 NO3−.
Table 6. Isotopic fractionation during NO3− uptake and NO3− reduction by Norin 8 and its nitrate-reductase deficient mutant (M819) supplied with NO3− (Experment 2)
Fractionation (Δ, ‰) during NO3− reduction (d − c)
a Cumulative N amounts supplied to the plants were 294 and 1176 mg per plant at 2 and 8 mol m−3 NO3− medium, respectively.
b Accumulation of plant N by NO3− N uptake during 30 d of NO3− feeding was calculated from the difference of plant N between 15- and 45-day-old plants. cData are mean ± SE of three replicates.
In short-term experiments, a higher N isotope fractionation was found for Kanto 168 than for Koshihikari. These results confirm the dependency of ɛ on substrate concentrations and plant species/cultivar, and thus help to explain the difference between previous observations. Evans et al. (1996) who used a low NH4+ concentration (50 mmol m−3) found almost no difference (+0·5‰) in δ 15N between the N of tomato plants and their medium N. Their tomatoes grew poorly in NH4+-N as opposed to when fed with low concentrations of NO3−-N. In contrast, Yoneyama et al. (1991) found differences in the δ15N values of rice plants and their medium, ranging from −4·1 to −17·5‰ and from −7·8 to –12·6‰, when fed with 1·4 and 7·2 mol m−3 NH4+-N, respectively. Even at similar low-to-medium NH4+ concentrations as supplied by Evans et al. (1996), isotopic fractionation during NH4+ uptake was observed by us for two contrasting rice cultivars (Tables 2 and 3). Significant ɛ values, up to −25‰, were observed during NH4+ uptake by marine bacteria and phytoplankton, when NH4+ concentrations were in the mmol m−3 range (Pennock et al. 1996; Waser et al. 1998). There is no conclusive evidence as to why the present results for NH4+-N nutrition (and those of others) differ from those of Evans et al. (1996). A taxon effect lies within reasonable possibilities. However, in view of the observed poor growth of tomato in NH4+-N nutrition in the report of Evans et al. (1996), carrier-mediated active transport may not be dominant in NH4+ uptake and instead, diffusion-mediated uptake could be a major path as a result of root damages in such plants. Nitrogen isotopic fractionation during 15NH4+ and 14NH4+ diffusion into the root cells at low NH4+ concentration could be smaller than that during active uptake. The NH4+ uptake mechanisms of poorly growing tomato should be examined.
As possibilities for isotopic fractionation during NH4+-N uptake, we suggest three processes. (1) Less competition between 15NH4+ and 14NH4 may occur during carrier-mediated plasma membrane transport at low concentrations (less than 200 mmol m−3 in this experiment) than at higher concentrations (here over 0·5 mol m−3). (2) Possibly there are higher concentrations of NH3, produced in the NH4+/NH3 equilibrium, in media with a high NH4+ concentration. NH3 produced in the equilibrium has lower δ15N values than the remaining NH4+ (Hermes, Weiss & Cleland 1985). At a medium pH of 6·5, as initially present in this experiment, the fraction of NH3 in the total of NH4+ + NH3 is very small (<0·01%), but because of the low δ15N value of NH3 and the rapid diffusion of NH3 through the membrane, in combination with an efficient assimilation of NH4+ by high-affinity glutamine synthetase that reduces the concentrations of NH4+ in the cytoplasm, the effect of NH3 uptake on ɛ values could be significant (see Hoch et al. 1992). (3) Some 15N-enriched NH4+ might be released from roots (Wang et al. 1993), after NH4+ assimilation by glutamine synthetase (Yoneyama & Kumazawa 1974; Yoneyama et al. 1993) at high NH4+ concentrations. Figure 1 shows the relationship between ɛ values and ƒ values (the percentage of total added N that remains in the medium). The ƒ values can be used as indirect measures of the NH4+ uptake rates, except for Koshihikari at high medium concentrations. As described before, ɛ values were lower at low-to-medium NH4+ concentrations when compared with high medium concentrations. Furthermore ɛ values were lower when ƒ values were low, i.e. when supplied N was nearing exhaustion. Both suggest that rapid NH4+ uptake may result in less N isotopic fractionation. With slow assimilation rates of NH4+ the quantity of effluxed 15N-enriched NH4+ might be larger, which results in higher ɛ values.
The small and concentration-independent N isotope fractionation that was found for NO3− uptake was expected, based on previous reports that described similar low ɛ values for different concentrations (Kohl & Shearer 1980; Mariotti et al. 1982; Bergersen et al. 1988; Yoneyama & Kaneko 1989; Evans et al. 1996). However, in contrast to NH4+ uptake, where discrimination always was in favour of net uptake of 14N-enriched NH4+ (Tables 2 and 3), NO3− uptake preference both for 15N and 14N was found on different occasions (Tables 3 5 and 6,). Kanto 168 showed a consistent preference for 14N (Table 5), as was observed before for Pennisetum (Mariotti et al. 1982). For Koshihikari discrimination was sometimes in favour of 14N, but mostly 15N was preferentially taken up and incorporated (Table 4), as was reported before for soybean (Bergersen et al. 1988). A higher net uptake of 15N-depleted NO3− may result from the efflux of unreduced 15N-enriched NO3− from the roots after absorption of NO3− and subsequent favouritism of 14NO3− over 15NO3− during NO3− reduction. Enrichment of plants with 15N may occur when 15N-depleted compounds are released after NO3− reduction in the plant (Robinson, Handley & Scrimgeour 1998). The fact that within one cultivar both types of discrimination were found (Table 4) indicates that both types of processing may occur simultaneously in plants, and that the net effect is the result of the balance between processes discriminating for 15N and processes discriminating for 14N. Which process dominates can depend on cultivar and N concentration in the substrate. For instance, at low and high concentrations Koshihikari and Kanto 168 showed opposite tendencies.
Although we used a NO3− reductase deficient mutant (M819), with an in vitro NADH-NR activity of less than 10% that of the parent Norin 8 (Hasegawa et al. 1992), no difference in ɛ values during whole-plant NO3− uptake was found. However, an increase of external NO3− concentrations from 2 to 8 mol m−3 resulted in equally negative ɛ values for NO3− uptake for both genotypes (Table 6). Unreduced NO3− as a fraction of the total amounts of absorbed NO3− was larger in M819 than in Norin 8 at 2 and 8 mol m−3 NO3−. These results indicate that NO3− reductase may not be involved in N isotope fractionation during NO3− uptake, but that extremely high NO3− concentrations in the medium may induce N isotope fractionation, independently of NO3− reductase activity, e.g. by osmotic stress or drought as reported by Handley et al. (1997) and Robinson et al. (2000). So far, no mechanism explaining N isotope fractionation by abiotic stresses is known.
It was concluded that in rice plants N isotope fractionation during NH4+ uptake is significant and depends on plant cultivar and available NH4+ concentrations. For NO3− uptake, only a small fractionation effect for whole plants was found over a wide range of NO3− concentrations, but at very high concentrations (8 mol m−3) a significant fractionation was found. N isotopic fractionation during whole-plant NO3− uptake did not depend on nitrate reductase activity.
We thank Dr H. Hasegawa for kindly supplying the rice mutant (M819).