Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 5 January 2006
Published in Agron J 98:168-176 (2006)
DOI: 10.2134/agronj2005.0180
© 2006 American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Agricola
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Related Collections
Right arrow Rice
Right arrow Crop Physiology & Metabolism
Right arrow Nitrogen
Right arrow Nutrient Management

Rice

Nitrogen Utilization Efficiency

Relationships with Grain Yield, Grain Protein, and Yield-Related Traits in Rice

Stanley Omar PB. Samontea,*, Lloyd T. Wilsona, James C. Medleya, Shannon R. M. Pinsonb, Anna M. McClungb and Joveno S. Lalesc

a Texas A&M Univ. System, Agric. Res. and Ext. Center, 1509 Aggie Dr., Beaumont, TX 77713
b USDA-ARS, 1509 Aggie Dr., Beaumont, TX 77713
c Dep. of Agronomy, Univ. of the Philippines atLos Baños, College, Laguna 4031, Philippines

* Corresponding author (sosamonte{at}aesrg.tamu.edu)

Received for publication July 1, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 NOTES
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Rice (Oryza sativa L.) breeders have historically not included nitrogen utilization efficiency (NUE) as a selection criterion in breeding for cultivars, even though it has economic and ecological implications. This study examined the significance and magnitude of variation in N content, NUE, N translocation ratio (NTR), and grain protein concentration among diverse rice genotypes. Fifteen rice genotypes were studied representing the combinations of low and high levels of four yield-related traits: maximum number of tillers, grain mass, main culm panicle node number, and panicle mass. These genotypes included ‘Lemont’, ‘Teqing’, and 13 advance recombinant inbred genotypes obtained from a Lemont x Teqing cross. Field data were obtained from experiments conducted during two cropping seasons. Plant samples were analyzed for N concentration. There was significant variation in N content and NUE among genotypes. The genotype NUE means ranged from 25.3 to 63.9 kg grain kg–1 N in a square meter of plants (kg grain kg–1 N), with the top four NUE values ranging from 56.6 to 63.9 kg grain kg–1 N. Nitrogen content and NUE were not significantly correlated with each other, and they had significant positive direct effects on grain yield. Grain yield was positively correlated with NUE, N content, and NTR, whereas NTR was correlated with grain protein concentration. Plant breeders could use these significant correlations to their advantage in breeding for rice cultivars that not only produce high yield but also utilize N efficiently and produce grain with a higher protein concentration.

Abbreviations: LSD, least significant difference • NTR, nitrogen translocation ratio • NUE, nitrogen utilization efficiency • TNC, total nonstructural carbohydrate


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 NOTES
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
RICE PRODUCTION in much of the world increasingly focuses on optimizing grain yield, reducing production costs, and minimizing pollution risks to the environment (Koutroubas and Ntanos, 2003). One of the inputs limiting rice production is N (Tirol-Padre et al., 1996). Nitrogen is essential to the rice plant, with about 75% of leaf N associated with chloroplasts, which are physiologically important in dry matter production through photosynthesis (Dalling, 1985). Rice plants require N during the vegetative stage to promote growth and tillering, which determines the potential number of panicles (Mae, 1997). Nitrogen contributes to spikelet production during the early panicle formation stage, and contributes to sink size by decreasing the number of degenerated spikelets and increasing hull size during the late panicle formation stage. Nitrogen contributes to carbohydrate accumulation in culms and leaf sheaths during the preheading stage and in grain during the grain-filling stage by being a component of photosynthesis (Mae, 1997).

Information on the seasonal patterns of N uptake and its partitioning within the crop is useful in assessing the amount, timing, and method of N fertilization to prevent the occurrence of N deficiencies, as well as to prevent overfertilization, which contributes to increased lodging, poor grain filling due to mutual shading, and increased severity and incidence of diseases (Liu, 1991; Saito, 1991). The development of efficient N management protocols requires recognizing cultivar differences and the critical stages of crop growth where fertilization is necessary to avoid potential yield loss (Senanayake et al., 1994).

Ladha and Reddy (2003) compared the rice grain yields and plant N requirements as they have increased through the years. Grain yields before the first green revolution were around 3 Mg ha–1, with the rice crop requiring 60 kg N ha–1. During the first green revolution, grain yields reached 8 Mg ha–1, with the rice crop requiring 160 kg N ha–1. The second green revolution is expected to produce grain yields of 12 Mg ha–1 and require 240 kg N ha–1.

Increasing the N concentration in rice plants does not always increase grain yield due to diminishing returns, and it is not always optimal from an economic perspective. The excessive use of N poses potential adverse environmental and health concerns (Bohlool et al., 1992), and increases incidence of foliar pathogens and plant lodging. Furthermore, the management of the N nutrition of the rice crop is difficult because lowland rice crop culture is conducive to N losses through ammonia volatilization, nitrification–denitrification, leaching, and runoff (Singh et al., 1998), which decreases the availability of N to the rice plant. The efficiency with which N is used by the rice plant is also affected by both N uptake efficiency and N utilization efficiency. The ratio of the amount of N uptake by the plant (or N content, kg N) over the amount of N supplied (kg Ns) is N uptake efficiency (kg N kg–1 Ns), while the grain yield (kg grain) to N uptake (kg N) ratio is N utilization efficiency (NUE, kg grain kg–1 N) (Moll et al., 1982). There are studies that recommend how to minimize N losses and improve N uptake efficiency in the lowland rice environment (Freney, 1997; Cassman et al., 1998; Raun and Johnson, 1999; Bautista et al., 2001; Singh et al., 2002).

The presence of significant variation in N utilization efficiency among rice genotypes that is related to grain yield is of particular interest to rice breeders. Moll et al. (1982) recommended selecting cultivars with both high N uptake efficiency and N utilization efficiency, while Liu (1991) recommended selecting appropriate cultivars to cope with the prevailing environmental and cultural conditions to improve efficient use of N. Previous studies have shown significant differences in NUE among rice genotypes grown at tropical (Tirol-Padre et al., 1996; Singh et al., 1998), subtropical (Ying et al., 1998), and Mediterranean environments (Koutroubas and Ntanos, 2003).

Grain protein concentration is directly related to the N concentration in the grain (Mossé, 1990). Children consuming high-protein (10%) milled rice showed improved growth compared with children consuming average-protein (6–7%) milled rice (Juliano, 1993). With rice being the most widely consumed cereal in the world (Obanni et al., 1998), it is important that rice breeders consider selecting genotypes with high efficiency in remobilizing N from vegetative parts to the grain (N translocation ratio, NTR, kg grain N kg–1 N) or genotypes with high grain protein concentration.

The objective of this study was to determine the significance and magnitude of variation in N content, NUE, NTR, and grain protein concentration among diverse rice genotypes. Specifically, we wanted to determine if there was a positive correlation between these four factors and grain yield.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
NOTES
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Field Experiment
Lemont, Teqing, and 13 rice genotypes (LQ170, LQ49, LQ55, LQ267, LQ82, LQ348, LQ199, LQ287, LQ125, LQ214, LQ81, LQ248, and LQ306 [Li et al., 1995]) derived from a Lemont x Teqing cross were used in this study. The rice genotypes represented the low and high level combinations of four important yield determinants: tillering ability, grain mass, main culm panicle node number, and panicle mass (Samonte et al., 1998). These growth parameters were selected based on population means of preliminary visual ratings (except grain mass, which was weighed) for each trait measured during the F8 generation. The LQ genotypes were F9 in 1994 and F10 in 1995. Lemont is a moderately high yielding, semidwarf, tropical japonica rice cultivar released in 1983, with good ratoon potential (Bollich et al., 1985). Teqing is a high-yielding, high-tillering, semidwarf indica Chinese rice cultivar (Wu et al., 1998).

The field experiments were conducted at the Texas A&M University System, Agricultural Research and Extension Center at Beaumont (29°57' N lat; 94°30' W long) during the 1994 and 1995 summer cropping seasons. The soil at this site is Beaumont, an Entic Pelludert (fine, montmorillonitic, and thermic), with a sand, silt, and clay composition of 3.2, 32.4, and 64.4%, respectively (Texas A&M University, 1971). Soil analyses estimated the pH at 6.0 and estimated the following nutrients in their available form: 1 g kg–1 (ppm) N, 4.2 g kg–1 P, 134.7 g kg–1 K, 3596 g kg–1 Ca, 480 g kg–1 Mg, 0.57 g kg–1 Zn, 62.7 g kg–1 Fe, 0.95 g kg–1 Cu, 132.2 g kg–1 Na, and 55.3 g kg–1 S. Each genotype was randomly planted in three plots in the rice paddy, for a total of 45 plots. Each plot was 6 m long and consisted of six rows, which were spaced 0.18 m apart. Rice was drill-seeded at the rate of 11.6 g m–2 on 20 Apr. 1994, and 15.4 g m–2 on 26 Apr. 1995. The plants were thinned to a uniform density of 20 seedlings per meter-row (112 seedlings m–2) when the seedlings reached the three-leaf stage and before permanent flood water was applied. In 1994, rice genotype LQ199 had low seed germination and was thinned to 15 seedlings per meter-row. Ammonium sulfate was applied by broadcasting at planting (34 kg N ha–1 in 1994 and 56.7 kg N ha–1 in 1995) and by topdressing after thinning (36 d after sowing, 79.4 kg N ha–1 in 1994 and 1995). In 1994, ammonium sulfate (79.4 kg N ha–1) was applied by topdressing at 67 d after planting due to leaf yellowing. The genotypes were topdressed with ammonium sulfate (79.4 kg N ha–1 in 1994 and 1995) on reaching their respective panicle differentiation stage. A higher amount of N was applied at planting in 1995 to avoid the leaf yellowing experienced in 1994. Permanent flood water was maintained at a 10-cm depth starting at fertilization following plant thinning.

Each plot contained five sampling quadrats, each measuring 0.72 by 0.30 m, in the inner four rows. During the sampling of a genotype, 1 of the 15 quadrats was randomly selected and all rice plants in that quadrat (about 24 plants) were dug out. This design enabled the space-efficient data sampling (destructive and nondestructive) of 15 genotypes. Sampling was conducted for each rice genotype about once every 12 d during the entire growth period for a total of six sampling dates for the earliest maturing genotype to 11 sampling dates for the latest maturing genotype. On each sampling date, plants were dug to a depth of 28 cm and washed to remove adhering soil. The tillers were counted and then divided into tiller groups: the main culms, the primary and secondary tillers, the tertiary and quaternary tillers, and a late tillers group. Within each tiller group, the plant samples were subdivided into stems (culms and leaf sheaths), leaf blades, and panicles. To minimize wilting and respiration during the separation of tillers and plant structures, the plant samples were kept in ice-cold water.

All samples were dried at 70°C in a ventilated oven after tillers and plant structures were separated. Dry masses were obtained after 48 h or until a constant mass was obtained. The panicle samples obtained during the last sampling date, which coincided with maturity, were hand threshed and the masses of the filled grain were used as estimates of grain yield. The samples were then ground using a cyclone sample mill (Udy Corp., Fort Collins, CO), with a mesh size of 0.5 mm. Dried and ground samples were sealed in plastic vials and stored at 4°C until analyzed for total N concentration.

Panicle node number, which is the node number of the main culm on which the panicle was borne, was obtained by counting the emergence of new leaves on 20 main culms for each rice genotype on a weekly basis. The heading date of each genotype was estimated as the number of days from sowing to the exsertion of 50% of the panicles in each plot. Harvest date was defined here as the date when about 80% of grain in the panicles were straw-colored and the grain in the lower portions of the panicle were in the hard-dough stage (De Datta, 1981). Minimum and maximum daily temperatures were obtained from the Texas A&M University System, Agricultural Research and Extension Center at Beaumont.

Nitrogen Determination
The procedure for N concentration determination in rice structural parts involved the potassium persulfate digestion method to convert organic N to NO3– (Purcell and King, 1996), the modified Cd-mediated reduction of NO3– to NO2– (Vodovotz, 1996), and the measurement of nitrite by the Griess assay (Vodovotz, 1996). The absorbance values were measured using an Elx800 microplate reader (Bio-Tek Instruments1) at 550 nm against a standard curve of NaNO2.

The persulfate digestion procedure described by Purcell and King (1996) was used in this study. For each sample, 0.05 g (±0.0001 g) of the dried, ground plant material was weighed and placed into a 50-mL centrifuge tube. Fifty mL of the digest solution, which contained 45 g K2S2O8 L–1 and 19 g NaOH L–1, was added and the tubes were autoclaved at 120°C (0.1 MPa) for 1.5 h. The autoclaved tubes were allowed to cool to ambient temperature, after which aliquots were taken from each tube for the Cd-mediated nitrate reduction procedure.

For each sample, 50 µL of the persulfate digest solution was transferred into a 1.5-mL centrifuge tube. Distilled H2O (750 µL) and ZnSO4 (40 µL, 30% w/v) were added to the tubes, the tubes were vortex mixed, incubated for 10 min, then centrifuged for 5 min. The supernatant (210 µL) of each sample was pipetted from the 1.5-mL tubes and transferred into fresh 1.5-mL centrifuge tubes containing one Cd granule (0.9–1.1 g). The tubes were then placed on a rotator (Glas-Col, Terre Haute) and rotated at about 55 rpm for 5 h. After rotation, the Cd granules were removed from the tubes and the tubes were centrifuged for 5 min.

Nitrite concentration was measured using the Griess assay. Three subsamples (50 µL each) of the resulting supernatant from each tube were placed into three adjacent wells of a 96-well microplate. A 50-µL solution containing 1% sulfanilamide and 2.5% H2PO4 was added into each well, and the mixture was allowed to incubate for 10 min. Afterward, 50 µL of a 0.1% N-(1-naphthyl) ethylenediamine dihydrochloride solution was added to form an azochromophore, and its absorbance was measured at 550 nm using an ELx-800 microtiter plate reader (Bio-Tek Instruments). To determine the total N concentration, the absorbance of each sample was compared with a standard curve produced by digesting three known amounts of KNO3 and a digest blank.

The Cd granules were recycled by placing 30 granules into a 50-mL centrifuge tube, adding 30 mL of H2O to each tube, and then rotating the tubes in a rotator (Glas-Col, Terre Haute1) at about 30 rpm for 15 min. Afterward, the water was drained and the procedure was repeated. After washing with water twice, 30 mL of 0.5 M HCl solution was added to each tube and the tubes were rotated for 15 min. The HCl wash was repeated twice. The final wash with 30 mL of 0.5 M NH4OH solution was also done twice.

Nitrogen Utilization Efficiency and Nitrogen Translocation Ratio
The N content (kg N) of a structural part was estimated as the mass of N in that structural part per 1 m2 ground area. Nitrogen utilization efficiency was computed for each genotype as the ratio of the mass of grain (kg grain) at harvest divided by the N content of plants (kg N) at heading. The N translocation ratio was computed for each genotype as the ratio of the mass of N in the grain (kg grain N) at harvest divided by the N content of plants (kg N) at heading. The N concentrations at heading obtained from the N determination procedure were used in the computation of N content, NUE, and NTR. When the field sampling date did not coincide with heading date, the N concentrations at heading were estimated through linear interpolation using the N concentrations obtained from samples taken immediately before and after the heading date.

Grain Protein Concentration
Grain protein content (g grain protein m–2) was estimated for each genotype in each year as the product of the mass of grain N (g grain N m–2) multiplied by the N/protein conversion ratio of k = 5.13 (Mossé, 1990). The grain protein concentration was estimated as the ratio of grain protein content per unit mass of grain (g grain protein kg–1 grain).

Statistical Analyses
Factorial analyses of variance were conducted to determine the significance of the effects of year, genotype, and tiller group on N content, NUE, NTR, and estimated grain protein concentration. The effect of the highest order interaction was assumed to be not significant and was used as the error term. Zar (1999) recommends this method and has found it to be robust when the highest interaction effect is not significant. When the highest interaction effect is significant, there is an increased chance of a type II error, that is, the opportunity to reject the null hypothesis correctly is lost and the F test increases in conservativeness. The year x genotype interaction was used as the error term for genotype. The means of significant main effects were analyzed using the least significant difference (LSD) mean comparison test.

Upon obtaining the F9 and F10 data during the 1994 and 1995 field study, the 15 rice genotypes were reclassified as belonging to either the high or low-level genotype group for the following traits: tillering ability, grain mass, main culm panicle node number, and panicle mass. The mean of the 15 genotypes for each trait was used as the basis for assigning genotypes into a high or low-level genotype group. Linear contrasts were performed to determine the effect of the low and high levels of these four genotypic categories on grain yield, NUE, NTR, and grain protein concentration.

Correlation analyses were conducted to determine the relationships among grain yield, maximum tiller density, 100-grain mass, main culm panicle node number, panicle mass, panicle density, N content at heading, NUE, NTR, and grain protein concentration. Hartley's test for homogeneity of variances was conducted for each of these traits to determine whether combining the 2 yr of data for correlation analyses was valid. The traits that were significantly correlated with grain yield (maximum tiller density, panicle mass, N content at heading, NTR, and NUE) were entered as predictor variables in a path analysis (Li, 1975; Williams et al., 1990) that estimated their direct effects (path coefficients, p) on grain yield.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 NOTES
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The rice genotypes used in this study were diverse (Table 1), as shown in the range of genotype means, averaged across years, for main culm panicle node number (13.6–19.9 nodes), maximum tiller density (406–790 tillers m–2), panicle mass (1.69–3.21 g panicle–1), 100-grain mass (1.70–2.73 g), and grain yield (150–372 g m–2). In this study, grain yields and N-related analyses (N content, NUE, NTR, and grain protein concentration) refer to data estimated from the main culms and primary and secondary tillers, which accounted for 85 ± 11% and 87 ± 7% of total grain yield in 1994 and 1995, respectively.


View this table:
[in this window]
[in a new window]
  		Table 1. Genotype means, averaged across years, for main culm panicle node number, maximum tiller density, days to heading, panicle mass, 100-grain mass, and grain yield of 15 rice genotypes grown at Beaumont, TX.

 
The error mean square values were relatively lower (either the lowest or second to the lowest) when compared with mean square values due to other sources of variation (year, genotype, tiller group, and their first-order interactions), as shown in the analysis of variance for N content, NUE, NTR, and grain protein concentration (Table 2). This supports the assumption that the highest interaction effect was not significant, adding credence to using the year x genotype x tiller group interaction effect as the error term.


View this table:
[in this window]
[in a new window]
  		Table 2. Analysis of variance of the effects of year, genotype, and tiller group on the N content, nitrogen utilization efficiency (NUE), nitrogen translocation ratio (NTR), and grain protein concentration (GPC) of 15 rice genotypes grown at Beaumont, TX, in 1994 and 1995.

 
Nitrogen Content and Nitrogen Utilization Efficiency
The factors that significantly affected the N content in rice plants at heading were year and genotype (Table 2). Total N content, averaged across genotypes and tiller groups, was greater during 1994 (5.97 g N m–2) than during 1995 (4.71 g N m–2). The genotype N contents, averaged across years and tiller groups, ranged from 3.51 g N m–2 for LQ287 to 8.37 g N m–2 for LQ348 (Table 3), showing significant variation among genotypes in their ability to take up N. Nitrogen contents of the combined main culm and first two tillers was 11.53 g m–2 at heading and 11.59 g N m–2 harvest in 1994. They were 10.19 and 8.86 g N m–2, respectively, in 1995. Since the mean N content was higher at heading (10.86 g m–2) than at harvest (10.22 g m–2), the N content at heading was used in the calculation of NUE. Guindo et al. (1994) reported a nonsignificant difference (LSD = 2.4 g m–2) between the plant total N content at heading (14.8 g N m–2) and at maturity (15.8 g N m–2).


View this table:
[in this window]
[in a new window]
  		Table 3. Genotype means (averaged across years and tiller groups) for N content and N utilization efficiency (NUE).

 
Rice genotypes should efficiently utilize the N taken up from the soil to produce grain. There was significant variation in NUE due to year, genotype, and year x genotype interaction (Table 2). The lack of a significant difference between NUE values obtained from the main culms and the first and second tillers indicated that the NUE is a genotypic parameter that can be estimated from samples obtained from either of these tiller groups. The mean NUE values (averaged across genotypes and tiller groups) were 50.9 kg grain kg–1 N in 1994 and 38.0 kg grain kg–1 N in 1995. There were lower plant dry masses and grain yields observed across Texas in 1995 and these were thought to have been caused by the high daily maximum temperatures during July and August 1995 (Wilson et al., 1996). In 1994 the mean maximum temperatures for July and August were 33.2 and 32.2°C, respectively, and in 1995 they were 33.7 and 34.0°C, respectively. Mean minimum temperatures in 1994 were 22.7°C for July and 21.9°C for August, and in 1995 they were 22.8°C for July and 23.4°C for August. Relatively higher temperatures (≥35°C) in 1995 than in 1994 most likely caused higher maintenance respiration (Yamamoto, 1954, as cited by Murata and Matsushima, 1975; Murata, 1964; Yoshida, 1981; Wilson et al., 1996), resulting in lower plant total nonstructural carbohydrate (TNC) concentrations and lower grain mass (Samonte etal., 2001), causing the lower NUE values in 1995 relative to 1994. Furthermore, the lower numbers of spikelets and filled grain may have contributed to the lower yields in 1995 than in 1994. Spikelets are prone to high temperature-induced sterility primarily at heading due to either poor pollen shedding or poor pollen germination and secondarily at around 9 d before heading corresponding to the young microspore stage (Satake and Yoshida, 1978).

The genotype NUE values (averaged across years and tiller groups) ranged from 25.3 to 63.9 kg grain kg–1 N, with the NUE of rice genotypes above the 75th percentile NUE value (LQ170, LQ199, LQ287, and LQ49) being significantly greater than NUE rice genotypes below the 25th percentile value (LQ306, LQ348, LQ267, LQ125) (Table 3). The NUE values of the four high NUE rice genotypes ranged from 56.6 to 63.9 kg grain kg–1 N, while the range of the low NUE rice genotypes was from 25.3 to 34.4 kg grain kg–1 N. Among the high NUE rice genotypes, LQ49 and LQ287 were the two earliest maturing genotypes and were the only genotypes that did not exhibit a decrease in stem TNC concentration when daily maximum temperatures exceeded 35°C (Samonte et al., 2001). Other studies (Ladha et al., 1998; Singh et al., 1998) reported significant differences in N content and NUE among genotypes. In a study conducted in Japan, the mean NUE of high yielding cultivars was 56.1 kg grain kg–1 N in 1993 and 58.1 kg grain kg–1 N in 1994 (Hasegawa, 2003). In a study conducted at China, Ying et al. (1998) reported that an N uptake of 250 kg N ha–1 and a NUE of 59 to 64 kg grain kg–1 N were required by cultivars that produced grain yields >13 Mg ha–1.

The ranking of genotypes based on NUE differed between years, but there was a significant trend of decreasing NUE as the number of days to heading increased (Fig. 1 ). This indicated that longer maturing rice had lower efficiency in using the N that it had taken up to produce grain.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between days to heading and nitrogen utilization efficiency (NUE) of 15 rice genotypes grown at Beaumont, TX, in 1994 and 1995.

 
The importance of NUE is more apparent when a comparison is made between genotypes with grain yields that are not significantly different. For example, LQ170 and LQ348 had grain yields (averaged across years and tiller groups) that were not significantly different (238.0 and 233.6 g m–2, respectively). However, their N contents (averaged across years and tiller groups) were significantly different at 3.69 g N m–2 for LQ170 and 8.37 g N m–2 for LQ348. The NUE values (averaged across years and tiller groups) of 63.9 kg grain kg–1 N for LQ170 and 28.9 kg grain kg–1 N for LQ348 indicate that LQ348 required 220% more N to produce the same amount of grain than LQ170.

Currently, rice breeders do not consider NUE in their selection and release of new cultivars. However, this comparison between the NUE values of two rice genotypes demonstrates the potential importance of selecting rice genotypes that have high NUE. Nitrogen utilization efficiency should be considered both when developing cultivars and when making fertilizer N recommendations.

Nitrogen Translocation Ratio
A rice genotype with a relatively higher NTR would be desirable as it would be more capable of remobilizing N into the grain during grain filling. Nitrogen translocation ratio was significantly affected by tiller group, year x genotype, and year x tiller group interactions (Table 2).

Nitrogen translocation ratio was significantly higher for the primary and secondary tiller group (0.49 kg grain N kg–1) than the main culm group (0.44 kg grain N kg–1). With regard to the significant year x tiller group interaction effect, the NTR of primary and secondary tiller group was significantly higher in 1994 than in 1995 and also significantly higher than NTR of main culms in both years.

Significant year x genotype interaction resulted in a different NTR ranking of genotypes between years. Genotype NTR ratios (averaged across tiller groups) ranged from 0.319 kg grain N kg–1 N for LQ306 to 0.689 kg grain N kg–1 N for LQ49 in 1994, and from 0.123 kg grain N kg–1 N for LQ348 to 0.761 kg grain N kg–1 N for Teqing in 1995. In comparison, Ying et al. (1998) estimated the NTR range of five cultivars to be 0.46 to 0.68 kg grain N kg–1 N at Los Baños, Philippines, and from 0.59 to 1.00 kg grain N kg–1 N at Yunnan, China.

In both years, there was a significant trend of decreasing NTR as the number of days to heading increased (Fig. 2 ). This indicated that longer maturing rice had lower efficiency in remobilizing the N that it had taken up into the grain.



View larger version (13K):
[in this window]
[in a new window]
  		Fig. 2. Relationship between days to heading and nitrogen translocation ratio (NTR) of 15 rice genotypes grown at Beaumont, TX, in 1994 and 1995.

 
Grain Protein Concentration
Grain protein concentration varied significantly due to year x genotype interaction effects (Table 2). Grain protein concentration (averaged across tiller groups) ranged from 34.5 g protein kg–1 grain for LQ170 to 70.0g protein kg–1 grain for LQ49 in 1994, and from 39.2g protein kg–1 grain for LQ55 to 90.0 g protein kg–1 grain for LQ267 in 1995. In 1994, two rice genotypes (LQ49 and LQ267) had grain protein concentration values >61.1 g protein kg–1 grain (the value halfway between the midpoint and the maximum value), whereas three rice genotypes (LQ287, LQ306, and LQ170) had N content values <43.3 g protein kg–1 grain (the value halfway between the midpoint and the minimum value). In 1995, two rice genotypes (LQ267 and Teqing) had grain protein concentration values >77.3 g protein kg–1 grain (the value halfway between the midpoint and the maximum value), while seven rice genotypes (Lemont, LQ306, LQ287, LQ82, LQ49, LQ348, and LQ55) had N content values <51.9 g protein kg–1 grain (the value halfway between the midpoint and the minimum value). This indicated that the distribution of grain protein concentrations of the rice genotypes used in this study was skewed toward the lower protein concentration levels. In comparison, based on a 5.13 N/protein conversion ratio, grain protein concentrations ranged from 65.2 to 106.2 g protein kg–1 grain (Mossé, 1990), while high protein milled rice had 82.1 g protein kg–1 grain (Juliano, 1993).

Contrast between Genotypes with Low and High Phenotypic Traits
Among four traits (tillering ability, grain mass, main culm panicle node number, and panicle mass), only panicle mass affected grain yield, with the genotypes that had heavier panicles (>2.15 g panicle–1) producing significantly higher grain yields (Table 4). None of the four traits had a significant effect on grain protein concentration. Both NUE and NTR were significantly affected by main culm panicle node number, with the genotypes that had more nodes (>16.7 nodes) having significantly lower NUE and NTR. In contrast, a simulation study by Wilson et al. (1998) showed that high main culm panicle node number improved grain yield, while a field study (Samonte, Wilson, and Tabien, unpublished research, 2004) showed that main culm panicle node number had positive effects on yield-related traits, such as LAI and the mass of panicles, leaves, and stems at heading. There is a need to study the dual effect of high main culm panicle node number, that is, its desirable positive effect of improving yield and yield-related traits, and its undesirable effect of lowering NUE and NTR during grain filling.


View this table:
[in this window]
[in a new window]
  		Table 4. Linear contrasts between the effects of low and high levels of tillering ability, grain mass, main culm panicle node number, and panicle mass on the grain yield, nitrogen utilization efficiency (NUE), nitrogen translocation ratio (NTR), and grain protein concentration (GPC) of 15 rice genotypes grown at Beaumont, TX, 1994 and 1995.{dagger}

 
Correlation among Traits
The variances of 10 traits (maximum tiller density, 100-grain mass, panicle node number, panicle mass, panicle density, grain yield, N content at heading, NUE, NTR, and grain protein concentration) were homogeneous across 1994 and 1995; hence the data from these 2 yr were combined for use in the correlation analyses. Grain yield was significantly correlated with panicle mass and maximum tiller density (Table 5). Nitrogen content was correlated with main culm panicle node number, as a plant with more nodes had more leaves and a greater number of potential tillers that require N. Main culm panicle node number was negatively correlated with NUE and NTR.


View this table:
[in this window]
[in a new window]
  		Table 5. Correlation matrix among 10 traits obtained from 15 rice genotypes grown at Beaumont, TX, in 1994 and 1995.

 
Nitrogen utilization efficiency was significantly correlated with grain yield (Fig. 3 ), panicle mass, panicle density, grain mass, and NTR (Table 5). Nitrogen content was significantly correlated with grain yield, panicle mass, and maximum tiller density. The path analysis results indicated that panicle mass, N content at heading, and NUE had significant positive direct effects on grain yield (Fig. 4 ). These implied that a genotype selected due to its high yield may be utilizing N efficiently or taking up high amounts of N. Because NUE and plant N content had a nonsignificant negative correlation, rice genotypes that absorb high quantities of N do not necessarily use their N efficiently to produce grain. It is logical that a rice breeder would select for a genotype that has both high yield and NUE, rather than select for high yield trait alone because of the possibility that the high yield is due to high N uptake. In a breeding program, the high yielding genotypes that are identified based on a yield trial can be screened for high NUE, so that the genotypes that are advanced for further testing or recommended as a cultivar are high yielders with high NUE. Singh et al. (1998) reported that N content was positively correlated with grain yield and negatively correlated with NUE.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 3. Relationship between nitrogen utilization efficiency (NUE) and grain yield of 15 rice genotypes grown at Beaumont, TX, in 1994 and 1995.

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 4. Path coefficients of the direct effects of selected traits (correlated with grain yield) on grain yield of 15 rice genotypes grown during the 1994 and 1995 cropping seasons at Beaumont, TX. Single arrowheaded lines represent path coefficients, while double arrowheaded lines represent significant correlation coefficients. MTD, maximum tiller density; PM, panicle mass; Nt, plant N content; NTR, nitrogen translocation ratio; NUE, nitrogen utilization efficiency.

 
There was a nonsignificant negative correlation between NUE and grain protein concentration, which supported a statement by Tirol-Padre et al. (1996) that the nutritional value of rice is not sacrificed with the improvement of NUE. However, Singh et al. (1998) reported a significant negative correlation between NUE and grain N concentration.

Nitrogen translocation ratio was significantly correlated with grain protein concentration and grain yield, while additional correlation analyses indicated that NTR was not significantly correlated with the N concentration of leaves, stems, or panicles at heading. In contrast, Singh et al. (1998) reported that NTR was significantly and negatively correlated with grain N concentration and grain yield in medium-duration rice, but not in long-duration rice.

There was a significant positive correlation (r = 0.45*) between days to heading and N content at heading, and a significant negative correlation (r = –0.68**) between NUE and days to heading. Furthermore, plant mass at heading had a significant positive correlation (r = 0.50**) with days to heading and a nonsignificant negative correlation (r = –0.28) with NUE. This indicated that genotypes that require more days to reach heading took up more N and were heavier at heading, but they did not necessarily use the N efficiently in producing grain. Tirol-Padre et al. (1996) reported that NUE was a more stable selection criterion than N content.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 NOTES
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
There was significant variation in N content and NUE among genotypes. Nitrogen content and NUE were not significantly correlated with each other, but both had significant positive direct effects on grain yield. Grain yield was correlated with NUE and NTR, whereas NTR was correlated with grain protein concentration. Plant breeders should take advantage of the significant variations and relationships among grain yield, NUE, NTR, and grain protein concentration. The high yielding genotypes that are identified through a yield trial can be screened for high NUE, NTR, or grain protein concentration, so that the selected genotypes are both high yielders and efficient utilizers of N. The space-efficient sampling method used in this study could be applied during this screening procedure.


   ACKNOWLEDGMENTS
 
This research was supported in part by funding from the Texas Rice Research Foundation and the Texas Agricultural Experiment Station provided to L.T. Wilson. The senior author appreciates the graduate fellowship provided to him by the Rockefeller Foundation, and the support of the University of the Philippines at Los Baños.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 NOTES
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
1 Reference to the manufacturer does not represent an endorsement over products by other manufacturers. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 NOTES
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Agricola
Right arrow Articles by Samonte, S. O. PB.
Right arrow Articles by Lales, J. S.
Related Collections
Right arrow Rice
Right arrow Crop Physiology & Metabolism
Right arrow Nitrogen
Right arrow Nutrient Management

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science 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