Nitrogen use efficiency in modern wheat cultivars

Bragantia, Campinas, v. 75, n. 3, p.351-361, 2016 ABSTRACT: The nitrogen use efficiency (NUE) is defined as the capacity of a given genotype in take advantage of the applied nitrogen (N) and transform it in biomass and grains. The objective of this study was to evaluate 12 wheat cultivars as to the NUE and its components. The experiment was conducted in a controlled environment, in a randomized block design with three replications. Twelve wheat cultivars were submitted to four N supply levels (0, 80, 160 and 240 kg of N∙ha–1). The data were submitted to analysis of variance, means multiple comparison, polynomial regression, and path analysis. The nitrogen remobilization efficiency (NRE) was the SOIL AND PLANT NUTRITION Article


INTRODUCTION
Nitrogen (N) is the most limiting nutrient for the production of wheat (Pan et al. 2006).Due to this fact and the possible environmental problems related to its use, the Nitrogen use efficiency (NUE) plays a fundamental role in sustainable grain production (Asplund et al. 2014).NUE is given by the ratio between grain yield (GY) and the amount of nutrient provided by the fertilizer (Moll et al. 1982;Cormier et al. 2013;Dai et al. 2013).
The mineral N fertilizer represents a significant cost in wheat production and may cause negative impacts on the environment through leaching and N 2 O emissions (Cui et al. 2014).Management practices to help farmers increase productivity and reduce production costs should be studied to ensure agricultural sustainability (Kaneko et al. 2010).In this sense, studies indicate that the development and use of wheat cultivars with higher NUE can contribute to reduce the applied N amounts without decreasing GY (Barraclough et al. 2014;Gaju et al. 2014).
Cultivars that use N more efficiently is one of the main objectives of wheat breeding programs (Sadras and Lemaire 2014).The main components of NUE are N uptake efficiency (NUpE), N utilization efficiency (NUtE), and N remobilization efficiency (NRE) (Le Gouis et al. 2000).The NUpE is the ability of plants to absorb the N available in soil.The NUtE is the relationship between crop yield and total N absorbed by the plant (N in grain + N in phytomass), indicating the GY obtained from each unit of N absorbed by the plant.The NRE is the ability of plants to translocate the N after anthesis from the shoot to the grains.Cultivars with higher NRE tend to accelerate the senescence process and increase N levels in grains (Gaju et al. 2014).
In wheat, the NUE is smaller than 60% (Haile et al. 2012;Hawkesford 2012;Duan et al. 2014).Rahman et al. (2011) indicated values between 28.8 and 40.0 kg grains per kgN applied that depended on the genotype and N levels effect, which ranged from 80 to 120 kg N•ha -1 .The variability of modern cultivars response to NUE has been attributed to NUpE (Sadras and Lemaire 2014), NUtE (Barraclough et al. 2010), and NRE (Kichey et al. 2007;Pask et al. 2012;Guo et al. 2014).The accumulation of phytomass (Giambalvo et al. 2010) and leaf chlorophyll content (Wani et al. 2011;Silva et al. 2014) are traits that have been linked to NUE and can, therefore, be applied for indirect selection of cultivars that use this nutrient more efficiently.
The first research stations investigating wheat crops date back to 1919 (Caierão et al. 2014) and were responsible for the development of pioneering cultivars, important for the Brazilian wheat.Beche et al. (2014) evaluated several Brazilian cultivars developed from 1940 to 2010 and observed that modern cultivars use N more efficiently and are more tolerant to low N availability compared to pioneer cultivars.In this study, we evaluated modern wheat cultivars regarding NUE and their components to establish the existence of genetic variability, useful for leveraging greater genetic progress in future breeding cycles.
Four N levels were evaluated: 0 (control), 3.7 (medium supply), 7.5 and 11.3 (higher supply) g N per pot.These levels represent, respectively, 0, 80, 160 and 240 kg N•ha -1 .The experiment was performed in a factorial (12 cultivars × 3 N levels + control) completely randomized block design with three replications.Each experimental unit consisted of two pots of 20 L (35 × 30 cm) with 30 homogeneous plants each.Table 1 shows soil physicochemical characteristics.The soil pH was corrected with dolomitic limestone (TNRP 85%) to obtain a base saturation value close to 70%.Similarly, the soil was corrected with 60 kg•ha -1 of potassium and phosphorus.
The N was applied in three phenological stages: I -1.4 g N per pot at the base for all treatments except control; II -65% of the remaining N at the start of tillering (Z 22; Zadoks et al. 1974); and III -35% at the end of tillering (Z 39; Zadoks et al. 1974).Urea (45% N), diluted with water, was the N source.The control of pests and diseases followed the recommendations for wheat crop.
Measurements of a*, b* and a + b* chlorophyll contents were held in Z 6.5 (Zadoks et al. 1974) of each cultivar, using the handheld ClorofiLOG CFL 1030-Falker.The readings were performed on the center of the flag leaf in 20 plants per experimental unit.
At the time of anthesis (Z 60), three plants per experimental unit were randomly collected to determine the N accumulated in the straw.At physiological maturity (Z 90), the other plants were harvested.The grain fractions and phytomass (shoots) were manually separated.The grain mass was measured, and the humidity, standardized to 13% to determine GY.The fractions of phytomass were dried at 40 °C for 48 h and grounded in a Wiley mill.Subsequently, samples of phytomass and kernels were subjected to chemical analysis in order to determine the N concentration following the Kjeldahl method (Tedesco et al. 1995).
Total N in the grains, in phytomass at anthesis a n d i n p hy s i o l o g i c a l m at u r it y w a s m e a s u re d by multiplying the N concentration (%) of the fraction by the phytomass production.The NUE measurements were calculated according to Moll et al. (1982), Guarda et al. (2004) and Foulkes et al. (2009): NUE (g•g −1 ) = GY/NS, where GY is the grain yield (g) and NS is the N supplied by the fertilizer (g); N Up E ( g • g − 1 ) = ( N G + N P M ) / N S , w h e r e N G is t he amount of N in g rains (g) and NPM is the amount of N in phytomass at physiological maturity (g); NUtE (g•g −1 ) = GY/(NG + NPM); and NRE (%) = NPM − (NPM − NG)/N anthesis , where N anthesis is the total NitrogenN at anthesis (g).
The data were tested for normal distribution (Kolmogorov-Smirnov test, p ≤ 0.05; Sprent and Smeeton 2007).Subsequently, the data were submitted to analysis of variance, considering the effects of cultivars, N levels and interaction as fixed.As a measure of experimental precision, selective accuracy was estimated (SA = √1 -1/Fc), following Resende and Duarte (2007), for the main effects of N and cultivars.The effect of N levels was measured by polynomial regression analysis tested up to cubic degree.Since no significant interaction was observed, the regression analysis was performed for the means of the cultivars, which were compared by Scott-Knott hierarchical clustering algorithm (p = 0.05) in general and within each N level (low and high N supply).Additionally, after checking the multicollinearity between the explanatory variables (Montgomery and Peck 1981), path analysis was carried out (Wright 1921) to identify the direct and indirect effects of the measured traits on GY and NUE.These analyses were performed by the Genes software (Cruz 2013).

RESULTS AND DISCUSSION
All traits were significant (p < 0.01) for the cultivar effect and N levels (Table 2).The NUpE, NUtE, NRE, N anthesis , NPM and N grains values were also affected by the cultivar versus N level interaction, indicating differences in the responses of the cultivars to the different N levels.The experimental precision related to the effects of N and cultivar, assessed by the magnitude of SA, is very high (SA > 0.90) according to the criterion established by Resende and Duarte (2007) for all traits.This precision favors the discrimination among cultivars.
To compare the means of the cultivars within the N levels (Table 3), these were grouped into two classes: low (between 0 and 80 kg N•ha -1 ) and high (between 160 and 240 kg N•ha -1 ) supply.Still, in this table, for the traits that showed no interaction, the means of cultivars were compared by the average of the four N levels.The NUpE, NUtE, NRE, N anthesis , NPM and N grains values differed among the cultivars for low and high N supply (Table 3), indicating the presence of several mechanisms responsible for the increase in NUE.Five cultivars (Topázio, BRS Parrudo, TBIO Iguaçu, TBIO Mestre, and TBIO Itaipu) were observed in the group with higher NUpE and low N supply.However, under high N supply, only one of these five cultivars (BRS Parrudo) remained in the group with the highest NUpE, showing efficient N uptake in both supply levels.NUpE had low amplitude variation among cultivars (1.54 to 1.89 g•g -1 for low N supply and 1.76 to 2.09 g•g -1 for high N supply), corroborating Haile et al. (2012).NUpE depends on the cultivar ability to recover the N applied.This possibly happened because the investigated cultivars are elite genotypes, with efficient N uptake.

pH H + Al
Regarding total N in phytomass at physiological maturity (NPM), only the Fcep Raízes cultivar remained in the best cultivar group for both N supply levels.Under high N supply, the BRS Gaivota and CD 150 are among the best genotypes and under low N supply; they represent intermediate cultivars for the NPM.The behavioral differences of cultivars in low and high N supply are also observed for NUtE, N anthesis and N grains .Overall, the cultivar BRS Parrudo had the highest NUpE, NUtE and N anthesis for high N supply and greater NUpE in low N supply.NUE was higher for the cultivar Mirante (58.55 g•g -1 ), followed by TBIO Itaipu (55.25 g•g -1 ), BRS Parrudo (52.73 g•g -1 ), and TBIO Iguaçu (52.24 g•g -1 ), and lower for the BRS Gaivota (37.59 g•g -1 ) (Table 3).The variation range of chlorophyll a* (Chl a : 33.6 to 37.5), b* (Chl b : 12.5 to 16.9) and total a + b* (Chl a+b : 46.1 to 54.4) indicated genetic variability for these traits.The highest chlorophyll levels were observed in BRS Parrudo (Chl a : 37.5; Chl b : 16.9; and Chl a+b : 54.4).The cultivar Mirante had high levels of chlorophyll (Chl a : 36.6;Chl b : 15.9; and Chl a+b : 52.6), the highest GY (74.3 g) and NUE (58.5 g•g -1 ).
The NUtE ranged between 25.64 and 38.91 g•g -1 for low N supply and between 18.86 and 28.70 g•g -1 for high N supply.The highest values were observed for the cultivars Mirante and TBIO Iguaçu (low N supply) and cultivars Mirante, TBIO Iguaçu, BRS Parrudo and TBIO Itaipu (high N supply); these cultivars also had high NUE (Table 3).
Table 4 shows the variation range of the results based on four N levels, per cultivar (traits with interaction) and overall.The data fitted polynomial regression models since the coefficients of determination (R 2 ) are high for each cultivar and low for overall.NUpE displayed increasing linear behavior for six cultivars and quadratic for four ones, with a maximum point within the limits studied.The critical point (CP, max) of the NUpE was observed at 222, 216, and 204 kg N•ha -1 for the cultivars BRS Parrudo,      4. Polynomial regression analysis for 11 plant traits of 12 modern wheat cultivars submitted to four Nitrogen levels (x, between 0 and 240 kg N•ha -1 ), per cultivar and overall.

Low N supply
NUpE = N uptake efficiency -g N in straw and grains per g N supplied (g•g -1 ); CP = Critical point (maximum or minimum); NUtE = N utilization efficiency -g grains per g N in the straw and grains (g•g -1 ); NRE = N remobilization efficiency (%); N anthesis = Total N at anthesis (%); NPM = Total N at physiological maturity (%); N grains = Total N remobilized to the grains (%); Chl a = Chlorophyll a (Falker Index); Chl b = Chlorophyll b (Falker Index); Chl a + b = Total chlorophyll; GY = Grain yield (g grains per plot); NUE = N use efficiency -g grains per g N supplied (g•g -1 ).
BRS Gralha Azul and BRS Gaivota, respectively.The NUpE decreased with increasing N supply for the cultivar BRS Tangará, while CD 150 and TBIO Iguaçu did not change.Guarda et al. (2004) and Asplund et al. (2014) also reported increasing NUpE for increasing N supply levels.
The NUtE decreased linearly for the cultivars BRS Gaivota and Fcep Cristalino.Also, it showed quadratic response with minimum point (at the levels indicated in the table) for BRS Tangará, TBIO Mestre, Mirante, BRS Gralha Azul, Topázio, CD 150, TBIO Iguaçu, TBIO Itaipu and Fcep Raízes and maximum point (N = 75 kg N•ha -1 ) for BRS Parrudo.The means of the cultivars showed a tendency to reduce the NUtE as the N supply level increased, stabilizing at 231 kg N•ha -1 (minimum CP).The differential behavior of cultivars with different N supply levels, regarding the NUtE, is important when defining the management and choosing the cultivar.
The NRE displayed a decreasing linear behavior for six cultivars and quadratic responses for the others.Maximum CPs were obtained for lower N supply for the cultivars BRS Tangará (41 kg N•ha -1 ), Mirante (12 kg N•ha -1 ) and TBIO Itaipu (4 kg N•ha -1 ).Furthermore, maximum CP was observed close to the highest N supply for the cultivars BRS Gralha Azul, BRS Gaivota and Topázio.The high N levels caused the N anthesis to increase, and the cultivars were unable to remobilize the N at the same rate when N supply was high compared to low supply.Barbottin et al. (2005) and Kichey et al. (2007) also reported that the NRE depends on the N level, and the highest remobilization rates occurred for lower N supply.
The N anthesis is the main source of N to the grains (Gaju et al. 2014).The highest protein concentrations in the grain are related to the higher remobilization of post-anthesis N (Barraclough et al. 2014;Bogard et al. 2010).According to Kichey et al. (2007), much of the N found in grains comes from the remobilization of N stored in the shoots and roots of the plant before anthesis.It is noteworthy that, as the N supply increases, most cultivars respond positively, demonstrating an average N anthesis increase of 0.0063% per kg of N applied.NPM and N grains also responded positively to an increasing N supply.For N grains , the means of the cultivars showed a quadratic response, with CP of 224 kg N•ha -1 (data not shown).
The direct and indirect effects of the seven traits on the NUE (Table 5) and GY (Table 6) were examinated by path analysis.It was observed a direct effect (DE) with sign and magnitude similar to the correlation coefficient (r) of the NRE over the NUE under low (DE = 0.94, r = 0.617) and high (DE = 0.65, r = 0.721) N supply, corroborating the studies of Barraclough et al. (2014), Guo et al. (2014) and Le Gouis et al. (2000).Although NUtE has been associated with the NUE (r = 0.487 and r = 0.516), this was mainly due to indirect effects (IE) of NRE (IE = 0.587) under low N supply.This result indicates that, for a group of modern cultivars, using the available N is linked to the ability of the cultivars to remobilize the nutrient toward the grains.In addition, under low N supply, the N anthesis affected directly the NUE (DE = 0.568) due to the DE of NREs with opposite sign canceling the correlation.This DE was not observed for high N supply (DE = 0.006).5. Direct and indirect effects of chlorophyll a, chlorophyll b, Nitrogen uptake efficiency, Nitrogen utilization efficiency, Nitrogen remobilization efficiency, total Nitrogen at anthesis and total Nitrogen at physiological maturity traits on Nitrogen use efficiency for low (0 and 80 kg N•ha -1 ) and high (160 and 240 kg N•ha -1 ) Nitrogen supply, Pearson correlation coefficient (r) and model coefficient of determination (R 2 ).

...continue
This is explained by the linear increase of N anthesis in most cultivars, with increasing N levels (Table 4); thus, high N supply results in high amount of N anthesis , and the RNE is presented as a measure of the highest association with NUE (DE = 0.654) (Table 5).
The association between N anthesis and GY results from the DE on the GY, under low (DE = 0.987) and high (DE = 0.930) N supply (Table 6).Accordingly, under low N supply, it is possible to select cultivars with greater NUE from the indirect selection on NRE or N anthesis .Under high N supply, N saturation probably occurs, and the NUE is then explained by the cultivars' ability to remobilize N from the straw to the grains, resulting in a higher grain quality and higher GY.Under low N supply, Beche et al. (2014) observed high DE of the NRE on GY while, under high N supply, they observed higher effect of NUtE and NUpE.This divergence in results might be explained by the different set of cultivars used in that study, characterized by pioneering and modern cultivars, whereas, in this study, we used only modern cultivars.
The Chl a , Chl b , and Chl a + b parameters are dependent on the cultivar (Table 3) and the N level applied (Table 4).However, the results in Tables 5 e 6 showed no important IE and DE on the NUE and GY.For low N supply, the correlation of GY with Chl a and Chl b is significant due to the IE of the N anthesis .Thus, in this study, the evaluation of Chl a , Chl b and Chl a + b was important to estimate the NUE.

CONCLUSION
The efficient use of Nitrogen by the evaluated cultivars resulted especially from the highly efficient way these cultivars were able to remobilize the absorbed Nitrogen to grain production.It is possible to select wheat cultivars with increased Nitrogen use efficiency from the indirect selection on Nitrogen remobilization efficiency or total Nitrogen at flowering.The Mirante, TBIO Itaipu, BRS Parrudo, and TBIO Iguaçu wheat cultivars are the most efficient in Nitrogen use, and the first two are the most efficient in Nitrogen remobilization.

Table 1 .
Chemical analysis results of the soil used in the experiment.

Table 2 .
Analysis of variance including mean square values of the effects of Nitrogen, cultivars, interaction and experimental error, with respective degrees of freedom, mean and selective accuracy for 11 plant traits of 12 modern wheat cultivars, under different N levels.

Table 3 .
Means for the traits of 12 modern wheat cultivars submitted to low (0 and 80 kg N•ha -1 ) and high (160 and 240 kg N•ha -1 ) N supply.