SEEDS ENRICHED WITH PHOSPHORUS AND MOLYBDENUM IMPROVE THE CONTRIBUTION OF BIOLOGICAL NITROGEN FIXATION TO COMMON BEAN AS ESTIMATED BY 15 N ISOTOPE DILUTION ( 1 )

Seeds with a high concentration of P or Mo can improve the growth and N accumulation of the common bean (Phaseolus vulgaris L.), but the effect of enriched seeds on biological N2 fixation has not been established yet. This study aimed to evaluate the effect of seeds enriched with P and Mo on growth and biological N2 fixation of the common bean by the 15N isotope dilution technique. An experiment was carried out in pots in a 2 x 3 x 2 x 2 factorial design in randomized blocks with four replications, comprising two levels of soil applied P (0 and 80 mg kg-1), three N sources (without N, inoculated with rhizobia, and mineral N), two seed P concentrations (low and high), and two seed Mo concentrations (low and high). Non-nodulating bean and sorghum were used as non-fixing crops. The substrate was 5.0 kg of a Red Latosol (Oxisol) previously enriched with 15N and mixed with 5.0 kg of sand. Plants were harvested 41 days after emergence. Seeds with high P concentration increased the growth and N in shoots, particularly in inoculated plants at lower applied P levels. Inoculated plants raised from high P seeds showed improved nodulation at both soil P levels. Higher soil P levels increased the percentage of N derived from the atmosphere (%Ndfa) in bean leaves. Inoculation with the selected strains increased the %Ndfa. High seed P increased the %Ndfa in inoculated plants at lower soil P levels. High seed Mo increased the %Ndfa at lower soil P levels in plants that did not receive inoculation or mineral N. It is concluded that high seed P concentration increases the growth, N accumulation


INTRODUCTION
The common bean (Phaseolus vulgaris L.) has been considered of low biological N 2 fixation capacity as compared to other grain legumes, which is attributed to the poor soil conditions where the crop is usually grown, the susceptibility of the species to nutritional and environmental stresses, the short vegetative period available for N 2 fixation, the sensitivity of the symbiosis to soil nitrate, and the competition from indigenous but often ineffective soil rhizobia (Piha & Munns, 1987;Leidi & Rodríguez-Navarro, 2000;Graham et al., 2003;Hungria et al., 2003).Nevertheless, ranges of biologically fixed N 2 in fieldgrown bean lines were reported as from 25 to 65 kg ha -1 (Ruschel et al., 1982) or 40 to 125 kg ha -1 (Rennie & Kemp, 1983) in shoots at physiological maturity, or from 21 to 44 kg ha -1 in grains (Pereira et al., 1989).Additionally, field experiments in seven countries showed that the N 2 fixation contributed with 35 % of the N accumulated by bean crops, with the highest figures at 70 % (Hardarson et al., 1993).Yet, using the 15 N natural abundance method, Kimura et al. (2004) estimated contributions of N 2 fixation ranging from 24 to 50 % in field-grown bean at different growth stages.Such evidence justifies efforts for improving the contribution of symbiotic N to the crop.
Low nutrient availability and soil acidity, typical of weathered soils, constitute serious restrictions to N 2 fixation in the common bean grown in the tropics (Graham et al., 2003).In acid soils, Mo is rapidly fixed to Fe and Al oxides and bound to organic matter, and the recovery from P fertilizers by crops is usually very low because most P becomes unavailable due to adsorption, precipitation or conversion to organic forms.The N 2 fixation under P deficiency is often reduced, which is usually explained by an effect of low P supply on the growth of the host plant, on the growth and functioning of the nodule, or on the growth of both plant and nodule (Almeida et al., 2000).Low P supply could induce changes in the relative growth of nodules and shoots rather than changes in leaf photosynthesis or N 2 -fixing activity per unit nodule mass (Hogh-Jensen et al., 2002).In addition to the P demand of the host plant, nodules require relatively higher amounts of P and energy than other plant tissues (Vadez et al., 1999;Olivera et al., 2004;Mortimer et al., 2008).Nodules of bean plants act as strong sinks of P and Mo to guarantee adequate rates of N 2 fixation (Brodrick & Giller, 1991;Almeida et al., 2000;Christiansen & Graham, 2002).
Seed constitutes a significant source of nutrients for seedlings and developing plants.Seed Mo content is sometimes enough for adequate growth during the whole growth season, and seed P content, although constituting a small proportion of total plant demand, may also be influential for plant establishment (Tyler & Zohlen, 1998).Moreover, nutrient-enriched seeds of leguminous plants can stimulate symbiotic N 2 fixation.Common bean plants raised from seeds with high P concentration showed increased growth, nodulation and N accumulation, particularly at low soil P levels (Teixeira et al., 1999;Araújo et al., 2002).Since the limited P supply retards the development of nodulation in the common bean (Araújo & Teixeira, 2000), P-enriched seeds can provide P during initial stages of nodule formation and growth (Thomson et al., 1991).Bean plants originating from seeds with high Mo concentration had higher accumulation of biomass and N in shoots and higher root nitrogenase activity (Brodrick & Giller, 1991;Kubota et al., 2008) and also produced more grains in soil with low N content particularly when plants did not receive additional foliar Mo (Vieira et al., 2005).In spite of such encouraging results, studies evaluating the simultaneous seed enrichment with P and Mo are not available.
Among the techniques suitable to quantify plantassociated N 2 fixation, the 15 N isotope dilution has been used widely for providing estimates of the proportion of N derived from N 2 fixation integrated over time (Unkovich et al., 2008).This technique involves the application of 15 N labeled fertilizer to the soil, and subsequent harvest of the N 2 -fixing and nonfixing control crops, which should both obtain N from the soil with the same 15 N enrichment (Unkovich et al., 2008).Hence, this work aimed to evaluate the effect of seeds enriched with P and Mo on the growth and biological N 2 fixation of common bean plants by the 15 N isotope dilution technique.

MATERIALS AND METHODS
Bean seeds of the Carioca cultivar enriched with P and Mo were obtained in a field experiment carried out at Santa Mônica Farm of Embrapa Gado de Leite (Valença -RJ), where different combinations of foliar fertilization with P and Mo were tested.At 52 and 71 days after seed emergence, 5 kg ha -1 of P as NH 4 H 2 PO 4 and 120 g ha -1 of Mo as (NH 4 ) 6 Mo 7 O 24 .4H 2 O were foliar sprayed.Plots also received foliar-applied urea to homogenize N levels.Seed samples suffered nitro-perchloric acid digestion, and in the extracts P concentration was measured colorimetrically by molybdenum-ascorbic acid, and Mo concentration by plasma emission spectrometry (ICP-EAS, Perkin-Elmer) (Table 1).
The experiment was conducted at Embrapa Agrobiologia (Seropédica -RJ), from October to December 2007.A 2 x 3 x 2 x 2 factorial scheme in randomized block design with four replications was used, comprising two levels of P applied to the soil (0 and 80 mg kg -1 of P), three N sources (control without N, seeds inoculated with rhizobia, and mineral N), two seed P concentrations (low and high) and two seed Mo concentrations (low and high).Non-nodulating bean genotype NORH-54 and sorghum cultivar BR-501 were used as non-fixing crops, which were also grown at two soil applied P levels (0 and 80 mg kg -1 ) with four replications.
The substrate was a 4 mm sieved Ap horizon of a clayey Red Latosol (Latossolo Vermelho) from Piracicaba -SP.The soil had received 10 kg ha -1 of N as ammonium sulphate with 10 atom % 15 N in excess in a field experiment conducted several years before, and was stored in sacks during the last few years.The soil presented 0.14 atom % 15 N in excess at the beginning of the experiment.Hence, interferences on the method due to spatial and temporal changes in 15 N enrichment of the soil mineral N during plant growth was rather reduced (Unkovich et al., 2008).Soil chemical analysis, performed as described by Embrapa (1997), showed: water pH 5.1, 0.0 mmol c dm -3 of Al 3+ , 42 mmol c dm -3 of Ca 2+ , 11 mmol c dm -3 of Mg 2+ , 28 mg dm -3 of available P (Mehlich-1), 121 mg dm -3 of available K. The soil was mixed with washed sand (1:1 in weight basis) and placed into 10 kg/pots.The following nutrients were applied diluted in water (in mg kg -1 of substrate): 10 Mg as MgSO 4 .7H 2 O, 2 Cu as CuSO 4 .5H 2 O, 1 Zn as ZnSO 4 .7H 2 O, 0.05 B as H 3 BO 3 , and 80 P as KH 2 PO 4 at high soil P level.Pots of low P level also received 75 mg kg -1 of K as KCl to uniform the K supply.In the treatments with mineral N, pots also received 60 mg kg -1 of N as urea.The substrate of each pot was then homogenized.At sowing, the substrate presented 12 and 49 mg dm -3 of available P (Mehlich-1), respectively, at the applied P levels of 0 and 80 mg kg -1 .
Seeds were sown six days after soil fertilization.In inoculated treatments, each seed received 1 mL of liquid inoculant containing the strains CIAT899 (or BR322) and PR-F81 (or BR520) of Rhizobium tropici from the collection of Embrapa Agrobiologia.Four bean plants were grown per pot after trimming.Pots were placed in open air, on tiles distributed on a greensward, and irrigation was performed daily.Twenty days after emergence, 300 mg N was applied per pot as urea in mineral N treatments.Abscised leaves were not collected during the experiment.Plants were harvested 41 days after emergence by the stage of pod setting.At harvest, shoots were cut at ground level, and leaves and stems were separately processed.Roots and nodules were recovered by washing the soil through a 2 mm sieve.Leaves, stems, roots, and nodules were oven dried, weighed and finely ground using a roll-mill.
Total N concentration was measured in each plant portion by the semi-micro Kjeldahl procedure.Accumulation of N was obtained by the product of N concentration and dry mass.The 15 N isotope composition was measured in one replication per treatment (except those that received mineral N) in leaves, stems and roots, using a continuous-flow isotope-ratio mass spectrometer (Finnigan DeltaPlus, Finnigan MAT, Bremen, Germany) at Embrapa Agrobiologia.Leaves and stems showed similar 15 N enrichment at both soil P levels (Table 2), and leaves corresponded to 54 % of the total N accumulated by the plants at harvest, averaged across all treatments.Therefore, the N isotopic composition was further measured only in leaves of bean plants under inoculation or without N, and in shoots of the nonnodulating bean and sorghum, of all replications.The percentage of N derived from the atmosphere (%Ndfa) was estimated by the formula (Unkovich et al., 2008): %Ndfa = [1 -(atom% 15 N excess of bean/atom % 15 N excess of non-fixing)] × 100 The analysis of variance was performed as a fourfactor design, evaluating the effects of soil P level, N source, seed P concentration, seed Mo concentration, and their interactions.The Tukey test at a 0.05 level was used to compare treatment means.

RESULTS AND DISCUSSION
The analysis of variance identified significant effects of the main factors of soil P level, N source and seed P concentration for shoot dry mass, nodule dry mass and N accumulation in shoots, whereas the effect of seed Mo concentration was significant for shoot N accumulation.Triple interaction between N source, seed P concentration and seed Mo concentration was significant for shoot mass, and the quadruple interaction was significant for nodule mass and for N accumulation in roots.Therefore, the means of each treatment are presented in figures to guide the interpretation of such complex effects, and the least significant difference calculated by the Tukey test at 0.05 level is presented to compare seed treatments within each soil P level and N source.

Plant growth
The highest P level applied to the soil strongly increased the shoot growth of bean plants, particularly in the control and inoculated plants (Figure 1a).Averaged across seed treatments, the highest soil P level increased shoot mass by 100 and 89 % in control and inoculated plants, respectively, whereas the highest P level increased shoot mass only by 16 % in plants with mineral N. It confirms that plants relying on symbiotic N had a larger demand for P to obtain an optimal growth than plants supplied with mineral N (Cassman et al., 1981;Israel, 1987).Plants under mineral N had greater shoot mass than the control and inoculated plants at both soil P levels, but the improved growth associated to mineral N was more intense at low soil P levels (Figure 1a).Control and inoculated plants yielded similar shoot mass.
At the lowest soil P level, inoculated plants raised from seeds with high-P high-Mo or with high-P low-Mo had greater shoot mass than plants from seeds with low-P low-Mo (Figure 1a).At the lowest soil P level, plants receiving mineral N originating from seeds with low-P high-Mo had lower shoot mass than those from seeds with high-P high-Mo.At the highest soil P level, control plants originating from seeds with Lowercase letters compare plant parts in columns (within bean tissue or non-fixing crops), and capital letters compare soil P levels in lines; means followed by the same letter did not differ by t test at 0.05 level.
high-P low-Mo had higher shoot mass than those from low-P low-Mo (Figure 1a).Therefore, the effect of seed P concentration on plant growth was more intense for inoculated plants at a lower soil P level.Teixeira et al. (1999) and Araújo et al. (2002) also observed that bean plants originating from P-enriched seeds showed improved shoot mass, but this effect was reduced as P level applied to the soil was increased.
The effects of Mo-enriched seeds on shoot growth were of small magnitude (Figure 1a).These results contrast with those of Kubota et al. (2008), which reported improved shoot growth of bean plants raised from seeds with high Mo concentration.
Nodule mass was strongly increased by a higher soil P supply (Figure 1b).Averaged across seed and N treatments, nodule mass increased by 101 % and shoot mass by 50 % when P was applied to the soil.It confirms that traits related to biological N 2 fixation are more responsive to soil P supply than plant growth (Almeida et al., 2000;Araújo & Teixeira, 2000).Even control plants showed a strong nodulation (Figure 1b), which reinforces the argument that the common bean is a promiscuous host regarding native soil rhizobia (Hungria et al., 2003).The addition of mineral N strongly reduced nodulation at both soil P levels (Figure 1b) since the Phaseolus-Rhizobium symbiosis is quite sensitive to soil nitrate (Westermann et al., 1981;Leidi & Rodríguez-Navarro, 2000).However, at the highest soil P level nodulation was less affected by the mineral N applied.Indeed, at the highest soil P level plants under mineral N had fewer but larger nodules, with an individual nodule mass of 2.79 mg/ nodule, as compared to 2.32 and 2.07 mg/nodule in the control and inoculated plants, respectively, averaged across seed treatments.Leidi & Rodríguez-Navarro (2000) observed that the increased P supplies improved nodulation of bean plants at low N concentrations in nutrient solutions, whereas the effect of P in high N concentrations was negligible.
Inoculated plants raised from high-P high-Mo seeds had higher nodule mass than plants raised from low-P low-Mo seeds at both soil P levels (Figure 1b).For control plants, high-P high-Mo seeds yielded more nodule mass than low-P high-Mo seeds at low soil P levels, whereas high-P low-Mo seeds yielded more nodule mass than the other seed treatments at high soil P levels (Figure 1b).Teixeira et al. (1999) verified that seeds with high P concentration enhanced nodulation of bean plants at every soil P level, but this effect was more intense at low soil P. High seed P concentration stimulated nodulation of lupin (Lupinus angustifolius L.) at every level of external P supply, but particularly under P deficiency (Thomson et al., 1991).
Seeds with low-P high-Mo concentration reduced nodule mass of plants that received inoculation or mineral N at the highest soil P level (Figure 1b).For control plants, low-P high-Mo seeds reduced nodule mass as compared to high-P high-Mo seeds at low soil P level, and as compared to high-P low-Mo seeds at high soil P level (Figure 1b).Kubota et al. (2008) also observed that bean plants originating from seeds with high Mo concentration showed reduced nodulation, although nitrogenase activity of root systems was not affected by seed Mo concentration.Since Mo supply is usually associated with increased nitrate reductase activity in bean plants (Pessoa et al., 2001), it is hypothesized that the additional Mo provided by enriched seeds would stimulate the assimilation of the N taken up from the soil, hence impairing nodule growth by a possible feedback mechanism induced by high concentration of reduced N compounds in roots (Leidi & Rodríguez-Navarro, 2000;Hogh-Jensen et al., 2002).

Accumulation of N
Plants that received mineral N showed similar N accumulation in shoots at both P levels, contrasting with control or inoculated plants where shoot N accumulation was higher when P was applied to the soil (Figure 2a).It demonstrates the higher P demand for N acquisition of plants depending on N 2 fixation (Israel, 1987).Seeds with high P concentration increased N accumulation in shoots of inoculated plants at the lowest soil P level (Figure 2a).High-P low-Mo seeds increased shoot N accumulation of control plants at the highest soil P level.Teixeira et al. (1999) also verified that bean plants originating from seeds with high P concentration had larger amounts of N accumulated in shoots.Seeds with high-P high-Mo increased N accumulation in the root systems of inoculated plants, as compared to low-P low-Mo seeds at low soil P, and as compared to low-P high-Mo seeds at high soil P level (Figure 2b).The highest N accumulation in the root system occurred in plants raised from high-P high-Mo seeds that received mineral N at the highest soil P level (Figure 2b), indicating a synergistic effect of enriched seeds and soil N and P supplies upon N acquisition.
Seeds with low-P high-Mo concentration reduced shoot N accumulation in plants that received mineral N at the lowest soil P level, and also reduced shoot and root N accumulation in inoculated plants at the highest soil P level (Figure 2a,b).Despite the reduced N accumulation in shoots of inoculated plants raised from low-P high-Mo seeds at the highest soil P level (Figure 2a), shoot mass was not reduced (Figure 1a), denoting an effect of seed Mo in diluting N in shoots.It could indicate a preferential effect of seed Mo on nitrate reductase activity, increasing N assimilation thus reducing plant demand for further N acquisition.Such hypothesis deserves further experimental work.

N isotope dilution
There was no significant effect of soil P levels on 15 N enrichment of non-fixing crops (Table 2).However, even using a homogenized soil supposedly stable in 15 N enrichment with time, non-nodulating bean and sorghum differed in terms of 15 N enrichment (Table 2).This reinforces the need to use different control crops to check the stability of soil 15 N labeling in order to provide more accurate estimates of N 2 fixation (Viera-Vargas et al., 1995).As non-nodulating beans best resemble the growth pattern of N 2 -fixing common beans, they were used as non-fixing crop in calculating the percentage of N derived from the atmosphere (%Ndfa) for the nodulated legume.The 15 N enrichment was similar in stems and leaves of bean plants, but roots showed higher 15 N enrichment (Table 2).It denotes that the N absorbed from the soil was accumulated rather by roots than shoots, whereas in shoots the N taken from the soil was diluted by symbiotically fixed N 2 .
Estimates of the %Ndfa ranged from 54 to 79 % (Table 3).Such figures are similar to those reported by Rondon et al. (2007), ranging from 50 to 72 %, but they are slightly higher than those reported by Viera-Vargas et al. (1995), varying between 49 and 61 %, for bean plants grown in pots with soil labeled with a single application of 15 N. Using the 15 N natural abundance, Mortimer et al. (2008) verified that %Ndfa in mycorrhizal beans grown in a nutrient solution  ranged from 50 to 55 % at 31 days after emergence.Hardarson et al. (1993) observed the highest values of %Ndfa in field-grown bean lines (70 %) when environmental conditions were favorable.
Inoculated plants had higher %Ndfa in leaves than control plants that did not receive inoculation or mineral N (Table 3).Therefore, in spite of the intense nodulation of control plants (Figure 1b), the inoculation of selected rhizobia strains contributed to improve N 2 fixation in common bean.Hungria et al. (2003) also verified that inoculation with selected strains increased N 2 fixation of field-grown common beans, expressed by higher N content in shoots and higher percentage of N as ureides.Higher soil P level increased the %Ndfa in bean leaves (Table 3).The 15 N isotope dilution technique used in field experiments also indicated that %Ndfa rises as soil P level increases in the common bean (Ssali & Keya, 1983) and lentil (Badarneh, 1995) crops.However, the %Ndfa in fieldgrown faba beans (Vicia faba) was not affected by P fertilization (Amanuel et al., 2000).
Seeds enriched with P increased the %Ndfa but this effect was significant only for inoculated plants at the lowest soil P level (Table 4).It indicates that seeds enriched with P, associated to inoculation of selected strains, can improve the contribution of N 2 fixation in the common bean, particularly in soils with low P availability.To our knowledge, it is the first report of enhanced contribution of N 2 fixation yielded by P-enriched seeds of legume plants.High-Mo seeds increased the %Ndfa but only in plants that did not receive N or inoculation at the lowest soil P level (Table 4).It is supposed that the higher %Ndfa in inoculated over control plants (Table 3) has reduced the additional effect of seed Mo on N 2 fixation of inoculated plants (Table 4).Yet in control plants, with lower diazotrophic efficacy of indigenous rhizobia strains, seed Mo would have stimulated N 2 fixation in a larger extent (Table 4).
These results demonstrate the beneficial effects of P-enriched seeds on improving growth and biological N 2 fixation of common bean plants.The effects of Mo-enriched seeds were less consistent than those reported by Vieira et al. (2005) and Kubota et al. (2008).Hence, the advantage of the simultaneous seed enrichment with P and Mo remains inconclusive.However, only one measurement at the stage of pod setting was performed in the present work, and other studies should be carried out in field conditions to ensure the effects of enriched seeds over the entire growth season.

CONCLUSIONS
1. Common bean plants accumulate more N and produce larger biomass of shoots and nodules when raised from seeds with high P concentration.
2. Seeds enriched with Mo increase the contribution of biological N 2 fixation in the common bean but only in plants that did not receive inoculation or mineral N grown at lower soil P levels.
3. Seeds enriched with P increase the accumulation of N in shoots and the contribution of biological N 2 fixation in the common bean, particularly for inoculated plants grown at lower soil P levels.

Figure
Figure 1.Shoot (a) and nodule (b) mass of common bean plants grown at two P levels applied to the soil (0 and 80 mg kg -1 ) and three N sources (CT control without N, IN inoculated with rhizobia, MN mineral N), originating from seeds with different concentrations of P and Mo (LP low P, HP high P, LM low Mo, HM high Mo); vertical bars represent the least significant difference by the Tukey test at 0.05 level, and compare seed treatments within each N source and soil P level.

Figure
Figure 2. of N in shoots (a) and roots (b, roots plus nodules) of common bean plants grown at two P levels applied to the soil (0 and 80 mg kg -1 ) and three N sources (CT control without N, IN inoculated with rhizobia, MN mineral N), originating from seeds with different concentrations of P and Mo (LP low P, HP high P, LM low Mo, HM high Mo); vertical bars represent the least significant difference by the Tukey test at 0.05 level, and compare seed treatments within each N source and soil P level.

Table 1 . Concentration of P and Mo, and 100-seed mass, of common bean seeds of the Carioca cultivar tested in the experiment
nd: values lower than 0.5 μg g -1 Mo, not detected by plasma spectrometry.

Table 3 . Contribution of biological N 2 fixation (in % of N derived from the atmosphere) in leaves of common bean plants at 41 days after emergence, grown at two levels of P applied to the soil (0 and 80 mg kg -1 ) and two N sources (control without N and inoculated with rhizobia); means of two P and two Mo concentrations in sowing seeds
Lowercase letters compare N sources in columns, and capital letters compare P levels in lines; means followed by the same letter, or without letters, did not differ by Tukey test at 0.05 level.

Table 4 . Contribution of biological N 2 fixation (in % of N derived from the atmosphere) in leaves of common bean plants at 41 days after emergence, grown at two levels of P applied to the soil (0 and 80 mg kg -1 ) and two N sources (control without N and inoculated with rhizobia), originating from seeds with low or high P and Mo concentrations
Lowercase letters compare seed P concentrations in columns, and capital letters compare seed Mo concentrations in lines; means followed by the same letter, or without letters, did not differ by the Tukey test at 0.05 level.