| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop, Soil, and Environmental Sci., 1366 W. Altheimer Drive, Univ. of Arkansas, Fayetteville, AR 72704
b International Crops Research Institute for the Semi-Arid Tropics, Patancheru 502 324 Andhra Pradesh, India
c USDA-ARS, Univ. of Florida, Agronomy Physiology and Genetics Laboratory, IFAS Building no. 350, 2005 SW 23rd Street, P.O. Box 110965, Gainesville, FL 32611-0965
* Corresponding author (lpurcell{at}uark.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: DOY, Day of Year • NDA, N derived from atmosphere • NN-Hardee, nonnodulating ‘Hardee’ • PNDA, proportion of N derived from atmosphere between harvests • RAU, relative abundance of ureides
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The proportion of a soybean crop's N derived from inorganic-soil N or from N2 fixation may also change depending upon whether or not the crop is exposed to drought or other environmental constraints. Field (Sinclair et al., 1987; Sall and Sinclair, 1991; Serraj et al., 1997) and greenhouse (Sall and Sinclair, 1991; Purcell et al., 1998) experiments have shown that N2 fixation in soybean is especially sensitive to water deficits and decreases before transpiration or photosynthesis. Furthermore, soybean grain yield under drought conditions was increased by 18% by applying 336 kg N ha–1 of NH4NO3 compared with treatments receiving no N fertilizer (Purcell and King, 1996). We are unaware of any soybean data that address how the proportion of N derived from N2 fixation or from the soil may change in response to drought.
Obstacles for measuring N2 fixation in field experiments include the labor required for collecting plant samples and the costs of 15N-enriched fertilizer and isotope analysis of plant samples. Typically, soybean and a reference crop that does not fix N2 are grown on soil enriched with 15N fertilizer, plant samples are harvested from an area of the plots, and the samples are dried, weighed, ground, and analyzed by mass spectroscopy. As the soybean crop fixes N2 from the atmosphere, 15N derived from the soil is diluted relative to that of the reference crop, and this relationship can be used to estimate the N derived from the atmosphere (NDA) (Fried and Middleboe, 1977; Herridge and Peoples, 1990):
 | [1] |
An alternative method for estimating NDA has been developed for soybean (Herridge and Peoples, 1990) and other legumes (Herridge and Peoples, 2002) that use allantoin and allantoate (collectively referred to as ureides) as their primary N-export product from nodules. In this procedure, the concentration of ureides from petiole-tissue extracts is expressed as the RAU, which is the fraction of the N from ureides relative to the sum of the N found in ureides and NO–3:
 | [2] |
The relationship between RAU and N2 fixation (as determined by 15N dilution) in soybean has been established in greenhouse-grown soybean plants using RAU values from both xylem sap and stem extracts from plants sampled during vegetative and reproductive development (Herridge and Peoples, 1990). By using this relationship between RAU and 15N dilution, Herridge et al. (1990) were able to predict accurately N2 fixation rates in field-grown soybean. This technique was also shown to be an effective tool in a breeding program for selecting lines capable of nodulating and fixing N2 in the presence of soil NO–3 (Herridge and Rose, 1994).
One possible source of error for the RAU technique that has not been evaluated is whether or not the relationship between RAU and N2 fixation changes under drought conditions. Because ureide concentration in xylem sap (Serraj and Sinclair, 1996) and stem extracts (de Silva et al., 1996; Serraj and Sinclair, 1996) increases greatly under drought conditions, it is important to establish the relationship between RAU and N2 fixation under water-limited conditions.
The objectives of our research were to: (i) evaluate the sensitivity of N accumulation from inorganic soil-N and of N2 fixation under well-watered and drought conditions in the field; (ii) compare the relationship between RAU and NDA from 15N dilution data for well-watered and drought conditions; and (iii) assess previous reports that N fertilization increased grain yield under drought conditions. These objectives were pursued in a series of field experiments in Florida, where we compared 15N-dilution and RAU techniques for estimating N2 fixation for well-watered and drought treatments, and in Arkansas, where we evaluated RAU and yield responses to N fertilization for well-watered and drought conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Nodulating and nonnodulating near isolines of ‘Hardee’ were sown on 4 Apr. 1996. Each plot consisted of seven rows, spaced 0.25 m apart that were 5.5 m in length. Nineteen and 20 d after sowing, plots were thinned to an intrarow spacing of 0.10 m, resulting in a population density of 40 plants m–2.
On 25 Apr. 1996, 25 mm of irrigation was applied to all plots at 0700 h. Beginning at 1100 h, appropriate amounts of 15NH4NO3 were applied in 2 L of water over the top of the crop canopy with a pressurized backpack sprayer to deliver 5, 50, or 100 kg N ha–1. The 15N atom excess for the respective N treatments was 10, 2, and 2%. Immediately after completing the N applications, all plots received an additional 8 mm of irrigation, which served to rinse fertilizer from leaves and move it into the rooting zone. On 2 and 17 May 1996, one-half the amount of 15NH4NO3 was applied (in 2 L of water) to the same plots as applied on the first N-application date. Methods of N application were similar to those described for the first date, but 12 mm of irrigation water was applied after N treatment.
Beginning on 8 May 1996, differential water treatments were begun for the well-watered and drought treatments (Table 1). On this date, plots of the well-watered treatment received 25 mm of irrigation water while there was no water applied to plots of the drought treatment. During the subsequent 20 d, well-watered plots received a total of 180 mm of rainfall and irrigation. Plots of the drought treatment received small amounts of irrigation when the first symptoms of water deficit were visible. The drought treatment received a total of 90 mm of irrigation and rainfall during this period. On 28 May 1996, there was 36 mm of rain, which terminated the drought treatment.
|
Total N in the finely ground subsample was determined by combustion analysis by the Soil Testing and Plant Analysis Laboratory at the University of Arkansas. The 15N concentration in the finely ground subsamples was determined by Isotope Services, Inc. (Los Alamos, NM). Samples were first combusted with a N analyzer (model NA 1500, Carlo-Erba, Milan, Italy), and combustion products were processed through a chemical train. The dinitrogen gas from the combusted sample was then passed through a VG Isotech Iomass Spectrometer (model Sira Series II, Cheshire, UK) that was calibrated with acetanilide standards and which quantified molecular mass peaks of 28, 29, and 30.
The NDA was determined by the 15N-dilution method (Eq. [1]; Fried and Middleboe, 1977; Herridge and Peoples, 1990). The 15N-dilution method determines N2 fixation by the dilution of N in plant tissues that occurs when plants fix their own N relative to that of a reference crop. In this case, the reference crop was the nonnodulated isoline of Hardee (NN-Hardee) at the same level of N application as the nodulated line. For each date when biomass was sampled, NDA was determined, and the proportion of NDA between harvests (PNDA) was calculated as (Herridge and Peoples, 1990):
 | [3] |
At four dates during the experimental period (Table 1), three petioles from fully expanded leaves at the top of the canopy were removed from each plot between 1100 and 1300 h. Petioles were dried at 80°C, bulked, finely chopped, and approximately 35 mg of dried tissue was placed in microfuge tubes. Ureides and NO–3 were extracted from petioles in 1.25 mL of 0.2 M NaOH at 100°C for 30 min (de Silva et al., 1996). Ureides and NO–3 in the extract were quantified using an autoanalyzer (model San System Plus, Skalar, Atlanta, GA) with the respective procedures described by Van Berkum and Sloger (1983) and Jackson et al. (1975). Relative abundance of ureide (RAU) was calculated from ureide and NO–3 concentrations as an estimate of the proportion of N derived from N2 fixation (Eq. [2]; Herridge and Peoples, 1990).
The average RAU for the four sampling dates of each plot of Hardee was used as an integrated estimate of the RAU between the two biomass-sampling dates. The average RAU was compared with estimates of PNDA from 15N-dilution measurements and to the calibrated relationship of RAU and 15N-dilution for vegetative soybean (Herridge and Peoples, 1990).
The experiment was a multiple split-plot arrangement of treatments in a randomized complete block design with four replications. The main plots were irrigation (well watered and drought stressed), subplots were three levels of N fertilization, and the final split was genotype.
Arkansas, 1995
Seeds of ‘Hutcheson’ were sown at a density of 28 m–2 on 22 May 1995 at the University of Arkansas Main Experiment Station in Fayetteville, AR (36°5' N, 94°7' W). The soil is a Captina silt loam (fine-silty, siliceous, active, mesic Typic Fragiudults).
Rows were 1-m apart and placed on raised beds that were 0.15 m high. Each plot consisted of eight rows that were 6 m in length. There were two yield rows, two sample rows, and four border rows per plot. The field had been previously cropped with soybean that was well nodulated, and no rhizobial inoculant was applied.
The experiment was a randomized complete block design with four replications and a split-plot arrangement of treatments. Irrigation was the main plot, and N treatments were subplots. All plots were well-watered with furrow irrigation until the R2 developmental stage (Fehr and Caviness, 1977) whenever the estimated soil-water deficit (Cahoon et al., 1990) reached 32 mm. After R2, the drought-stressed treatments were dependent solely on rainfall.
Within each main plot, four N treatments were established: (i) no N fertilizer applied, (ii) 112 kg N ha–1 applied at V6, (iii) 112 kg N ha–1 applied at R2, (iv) 112 kg N ha–1 applied at V6 and an additional 112 kg N ha–1 applied at R2. This last treatment will be referred to subsequently as the V6&R2 treatment. Ammonium nitrate fertilizer was used as the N-fertilizer source. The appropriate amount of fertilizer was weighed for each row, and this was distributed evenly by hand on the soil surface. Immediately after N application, 12 mm of water was applied to all treatments with sprinkle irrigation.
Three petioles from mature leaves at the top of the canopy were removed from each plot approximately every 2 wk during the season, and they were dried and analyzed for ureide and NO–3 using the same procedures described for the Florida experiment. At maturity, seed yield was determined by combining two end-trimmed rows that were 4.2 m in length.
Arkansas, 1996
The experiment conducted in 1995 in Fayetteville was repeated in 1996 on a similar soil. For 1996, the sowing date was 7 June 1996. Experimental design was similar to the 1995 experiment except that a randomized complete block design with a factorial arrangement of water and N treatments was used rather than a split-plot arrangement of treatments. The irrigation method in 1996 was drip irrigation.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The rate of N accumulation between the two biomass sampling dates had significant interactions among factors of irrigation, N treatment, and genotype. For NN-Hardee, N accumulation was similar between drought and well-watered treatments within each level of N treatment (Fig. 1) . For the drought and well-watered treatments of NN-Hardee, N accumulation rate increased approximately eight-fold as the N application was increased from 10 to 200 kg N ha–1.
|
Analysis of variance of petiole NO–3, ureides, and RAU for the four sampling dates indicated a complex response to the main effects of irrigation, N, and genotype and to their two-way and three-way interactions (Table 2). The three-way interaction was significant on 16 May for NO–3, 13 and 16 May for ureide, and 9 and 20 May for RAU. For measurements made on 9 and 13 May, if the three-way interaction for these variables was nonsignificant, then one or more of the two-way interactions were significant. Conversely, on 16 and 20 May, if the three-way interaction was nonsignificant, then the two-way interactions were also nonsignificant, but one or more of the main effects were significant. To simplify data presentation and to provide a complete profile response of petiole NO–3, ureides, and RAU to irrigation, N, and genotype, we have chosen to present means of the three-way interactions (Table 3).
|
|
Petiole ureide concentration was generally not affected by N treatment in either genotype, or it was higher for the low-N treatment than for the medium- and high-N treatments (Table 3). The petiole ureide concentration for the low-N treatment of Hardee was particularly high on 9 May for the well-watered treatment in which it was three-fold greater than the ureide concentration for plants of the medium- and high-N treatments.
Petiole ureide concentration of Hardee increased on 13 and 16 May in response to drought (Table 3), and these dates corresponded to the dates with the most severe stress. The accumulation of ureides in response to drought has been noted in previous reports (de Silva et al., 1996; Serraj and Sinclair, 1996). Despite a complete lack of nodules, NN-Hardee had measurable ureides at all dates. Although the ureide concentration was generally lower in NN-Hardee than in Hardee, ureide concentration tended to increase for the high-N plots of the drought treatment.
Petiole NO–3 and ureide concentrations were used to calculate RAU, and differences in RAU among treatments were generally dominated by the effect of N treatment on petiole NO–3 (Table 3). Increased amounts of applied N resulted in decreased RAU. The decrease in petiole NO–3 concentration for Hardee at the two later sampling dates, particularly for the well-watered treatment, resulted in higher RAU values than those observed at the two earlier dates. Despite the fact that NN-Hardee did not have nodules and did not produce ureides from N2 fixation, RAU values for NN-Hardee were generally similar to the values found for Hardee for the same irrigation and N treatments.
At the same level of N treatment, petiole NO–3 concentration was generally higher for the drought treatment than for the irrigated treatment, which may have been due to less NO–3 leaching from plots of the drought treatment, or to a decreased assimilation and utilization of NO–3 for the drought treatment, or to a combination of these two factors. The parallel increases in both NO–3 and ureide concentrations for the drought treatment tended to offset one another such that there were generally not large differences in RAU for irrigated and drought treatments within a N treatment on a given date.
The RAU values were calculated (Eq. [2]) from ureide and NO–3 data from each experimental unit and then averaged across replications for presentation in Table 3. Consequently, values of RAU in Table 3 may differ from calculations of RAU using the mean values of ureide and NO–3. This difference was particularly evident for the low-N treatments, in which the NO–3 concentration was 0 for some observations and >0 for other observations of the same treatment.
The PNDA for the irrigated treatment, calculated using the 15N-dilution method, ranged from approximately 0.4 for the high-N treatment to 0.95 for the low-N treatment (Fig. 2) . For the drought treatment, the PNDA covered a much greater range: from approximately 0.0 for the high-N treatment to 0.9 for the low-N treatment (Fig. 2). Although the total N accumulation rate for the high-N treatment under drought was approximately the same as for the high-N, well-watered treatment (Fig. 1), <25% of this N was from N2 fixation (Fig. 2). Therefore, the greater N accumulation rates for the high-N treatment under drought shown in Fig. 1 were due to uptake and assimilation of inorganic N rather than stimulation of N2 fixation.
|
|
Arkansas, 1995
From Day of Year (DOY) 210 to 243, total rainfall was 4.5 mm (Fig. 3A)
and maximum temperatures averaged 34°C, which resulted in substantial drought stress during the podfill stages. During peak stress, the soil water potential at a depth of 15 cm was <–80 kPa (data not shown).
|
|
|
Arkansas, 1996
In general, 1996 was a more favorable growing season than 1995. From sowing to R6, rainfall totals were 396 mm in 1995 and 443 mm in 1996. Moreover, rainfall was more evenly distributed in 1996 than in 1995 (Fig. 3A and 3B), and drought stress was probably less severe.
Relative abundance of ureide in petioles generally increased throughout the 1996 season, similar to the 1995 season (Fig. 4B). After the initial application of N at V6, there were significant main effects of N treatment on RAU at subsequent sampling dates. There was no significant interaction between irrigation and N treatment on RAU, however, and the response of RAU to N treatment has been averaged across irrigation treatments (Fig. 4B). Application of N fertilizer at V6 and at R2 delayed the increase in RAU, indicating that for these treatments, the proportion of N derived from N2 fixation was delayed (Herridge and Peoples, 1990). When plants were sampled on DOY 248 (90 d after planting), there was also a significant main effect of irrigation with drought treatments having less RAU (0.62) than the well-watered treatments (0.72). This difference in RAU might reflect a higher N2 fixation rate for the well-watered treatment. Alternatively, the higher RAU for the irrigated treatment may be due to a greater depletion of N fertilizer from the soil due to crop utilization and/or leaching.
Although there was no significant interaction between irrigation and N treatments for yield, there were appreciable main effects of N and irrigation treatments (Table 6). The main effect of irrigation on yield was significant with well-watered treatment having 9% greater yield than the drought treatment. Averaged across irrigation treatments, N treatment at V6&R2 gave higher yields compared with the control treatment receiving no N fertilizer; however, in 1996 there was no yield response to N treatment applied only at R2, unlike the 1995 results.
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because ureides are generally assumed to be closely linked to N2 fixation, it is somewhat surprising that RAU values were similar between nodulating and nonnodulating isolines of Hardee at Gainesville (Table 3). Nonnodulated bean (Phaseolus vulgaris L.) also contained relatively high concentrations of ureides in leaf and stem tissues (Thomas et al., 1980), and inhibitor experiments indicated that the likely source of ureides was from purine catabolism. The enzyme that breaks down allantoate, allantoate amidohydrolase, requires Mn2+ as a cofactor, and in many soybean cultivars, Mn2+ appears to be limiting or is unavailable for optimum ureide breakdown (Purcell et al., 2000). Although the NN-Hardee was completely dependent upon soil N, ureides in petiole tissues were readily detectable. Furthermore, the increase in petiole ureides that occurs in response to drought stress (de Silva et al., 1996; Serraj and Sinclair, 1996) was also observed in NN-Hardee for the high-N treatment (Table 3). These results indicate that caution should be exercised when using RAU as a proxy for N2 fixation in soybean genotypes or in Bradyrhizobium japonicum strains that have different efficiencies of N2 fixation. Our RAU data for the irrigated treatment agreed within 15% of the published calibration of RAU and PNDA from 15N dilution (Herridge and Peoples, 1990). For the drought treatment, however, there were large differences between the relationship of RAU and PNDA that we observed and with the published calibration. Further work is required to define the interactions of RAU with symbiotic effectiveness and with drought stress before this approach can be broadly applied for quantifying N2 fixation under drought conditions.
For the Fayetteville location in both years, RAU was <0.20 during vegetative development in the absence of N fertilization (Fig. 4A, 4B). In contrast, RAU in the Gainesville experiment was >0.80 for vegetative Hardee plants (Table 3), indicating a greater dependence on N2 fixation at Gainesville than at Fayetteville (Herridge and Peoples, 1990; Fig. 2), which is not unexpected given the low soil organic matter and propensity for leaching on the sandy soil in Florida.
A greater N accumulation rate for N-fertilized treatments may have resulted in the increased yield that was observed in the experiments conducted in Arkansas. In 1995, N fertilization increased yield 15 to 25% in drought treatments and 12 to 15% in the well-watered treatment. In 1996, averaged across irrigation treatments, N fertilization at V6&R2 increased yield 10% more than the control treatment. Previous reports on yield responses to N fertilization when water supply was adequate have been conflicting. In some cases, relatively low rates of N fertilizer (50 to 100 kg N ha–1, Yinbo et al., 1997) have resulted in large yield increases (64%), but in other cases low rates of N fertilizer (35 kg N ha–1, Heatherly et al., 2003) have had no effect on yield. Likewise, high rates of N fertilizer have had conflicting results. Brevedan et al. (1978) found yield increases of approximately 800 kg ha–1 when 168 kg N ha–1 was applied at R2, whereas Weber (1966) found no yield response in a nodulating soybean line with fertilizer application as high as 672 kg N ha–1.
There are several possible reasons for the different yield responses to N fertilizer in the absence of obvious drought stress that have been reported. The first is that water-deficit stress is not easily quantified, and a mild water deficit, sufficient to inhibit N2 fixation, may not be readily recognized. Under these conditions, a greater tolerance to drought of uptake and assimilation of inorganic N would restore crop growth and increase yield relative to that of a crop dependent primarily upon N2 fixation. A second possible explanation for different responses of yield to N fertilizer under well-watered conditions is that the soybean crop may have been nodulated by strains of B. japonicum with different symbiotic efficiencies. For inefficient (but highly competitive) strains, supplemental N would likely benefit the crop regardless of irrigation treatment. Responses of soybean yield to N fertilizer may also differ, depending upon soil characteristics including residual soil N and soil organic matter.
This research confirmed previous reports that N2 fixation in soybean was more sensitive to drought stress than was the uptake and assimilation of soil N and that fertilizing soybean with large amounts of N fertilizer could increase yield under drought conditions (Purcell and King, 1996). Although application of high rates of N fertilizer may not be an economical means of increasing drought tolerance in soybean, it illustrates the potential gain in yield that perhaps could be achieved if the tolerance of N2 fixation to drought was genetically enhanced to reach the same level as that of the uptake and assimilation of soil N. In that genotypes have been discovered with increased tolerance of N2 fixation to drought (Sall and Sinclair, 1991; Sinclair et al., 2000; Purcell and Specht, 2004), selection for drought-tolerant N2 fixation appears to be an important avenue for future cultivar improvement.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication July 3, 2004.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. A. King and L. C. Purcell Genotypic Variation for Shoot N Concentration and Response to Water Deficits in Soybean Crop Sci., October 2, 2006; 46(6): 2396 - 2402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Ray, L. G. Heatherly, and F. B. Fritschi Influence of Large Amounts of Nitrogen on Nonirrigated and Irrigated Soybean Crop Sci., December 2, 2005; 46(1): 52 - 60. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
