Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 March 2007
Published in Crop Sci 47:519-528 (2007)
© 2007 Crop Science Society of America
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 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Agricola
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Related Collections
Right arrow Nitrogen
Right arrow Maize
Right arrow Crop Genetics

CROP BREEDING & GENETICS

Nitrogen Uptake and Utilization in Contrasting Nitrogen Efficient Tropical Maize Hybrids

Mosisa Workua, Marianne Bänzigerb,*, Gunda Schulte aufm Erleyc, Dennis Friesend, Alpha O. Diallob and Walter J. Horstc

a Bako Agricultural Research Center, P.O. Box 03, Bako, West Shoa, Oromia, Ethiopia
b CIMMYT-Kenya, P.O. Box 25171, Nairobi, Kenya
c Institute of Plant Nutrition, Univ. of Hanover, Herrenhaeuser Str. 2, D 30419 Hanover, Germany
d CIMMYT-Ethiopia, P.O. Box 5689, Addis Ababa, Ethiopia

* Corresponding author (m.banziger{at}cgiar.org).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Maize cultivars with improved grain yields under nitrogen (N) stress are desirable for sub-Saharan African maize growing environments. This study assesses N uptake, N utilization, and the genotype x environment (G x E) interaction of 16 tropical maize (Zea mays L.) hybrids differing in grain yield under low-N conditions. Hybrids were evaluated under low-N, medium-N, and high-N at Harare, Zimbabwe, in 2003 and 2004 and at Kiboko, Kenya, in 2003. At maturity, N accumulation in the aboveground biomass ranged from 47 to 278 kg N ha–1 in various experiments. Grain yields ranged from 1.5 to 4.3 Mg ha–1 and 10.6 to 14.9 Mg ha–1 for the same experiments, respectively. Significant G x E interactions were observed which became more pronounced as the difference in N stress intensity between two environments increased. High grain yield under low-N was consistently associated with higher postanthesis N uptake, increased grain production per unit N accumulated, and an improved N harvest index. Additive main effect and multiplicative interaction analysis identified hybrids with specific adaptation to either low-N or high-N environments. Several hybrids produced high yields under both low-N and high-N conditions. More detailed studies with these hybrids are required to examine the underlying physiological mechanisms contributing to the N-use efficiency.

Abbreviations: AD, anthesis date • AMMI, additive main effect and multiplicative interaction • G x E, genotype x environment • GY, grain yield • H, hybrid • IPCA1 and IPCA2, axes 1 and 2 of the interaction principal component analysis • LSD, least significant difference • N, nitrogen • NUA, nitrogen uptake at anthesis • NUAP, nitrogen uptake between anthesis and physiological maturity • NUT, N utilization • NHI, N harvest index • STONP, stover N at physiological maturity • Z03N1, Z03N2, Z03N3, K03N1, K03N2, K03N3, Z04N1, Z04N2, Z04N3 are abbreviations for experiments coded for location (K—Kenya, Z—Zimbabwe), year (2003 or 2004), and N level (N1—low-N, N2—medium-N, N3—high-N).


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ENVIRONMENT is a general term that covers conditions under which plants grow and may vary between locations and years with respect to climatic variables (radiation, temperature, rainfall), soil fertility levels, other management practices, or combinations of these factors (Romagosa and Fox, 1993). Every nongenetic factor that is a part of the plant environment has the potential to cause differential genotypic performance resulting in a significant G x E interaction (Fehr, 1991). In sub-Saharan Africa, low soil fertility, especially low N is among the major abiotic stresses limiting maize production in fields of smallholder farmers. In most cases, less than 20 kg N ha–1 is applied to maize crops because farmers lack access to fertilizer or do not have the cash to buy this input (Sibale and Smith, 1997; Friesen et al., 2002). Studies on a large number of maize genotypes found that G x E interactions become more important in maize as N stress increases (Bänziger et al., 1997; Presterl et al., 2003), indicating the need for maize breeding programs which specifically target the low-N conditions of farmers in sub-Saharan Africa.

A wide range of researchers have reported the existence of genetic variability in N efficiency under low-N conditions (e.g., Lafitte and Edmeades, 1994; van Beem and Smith, 1997; Below et al., 1997; Akintoye et al., 1999; Bänziger et al., 1999; Horst et al., 2003; Presterl et al., 2003). Efficiency is the ability of a system to convert inputs into desired outputs or to minimize the conversion of inputs into waste (Lynch, 1998). The amount of mineral nutrients might be considered as an input, whereas plant growth, physiological activity, or yields are typical outputs. Nitrogen efficiency has been defined as the ability of a genotype to realize an above average grain yield under conditions of low N availability or suboptimal N supply (Graham, 1984; Sattelmacher et al., 1994).

There are several studies that evaluated the relative importance of N uptake and utilization of accumulated N for grain yield production. Moll et al. (1982) evaluated eight unselected single cross hybrids under low-N and high-N levels and attributed genetic differences in grain yield under low-N to differences in N utilization. Under high-N, they attributed genetic differences in grain yield largely to variation in N uptake. Kamprath et al. (1982) also compared three population hybrids under three N levels. They associated higher grain yield under low-N and medium-N with a higher N uptake, whereas higher grain yield under high-N was linked to the ability to utilize N accumulated in the plant. Other authors reported that high N efficiency was achieved by a combination of high N uptake and N utilization in maize (Wiesler et al., 2001) and in wheat (Triticum aestivum L.) (Ortiz-Monasterio et al., 2001). Variation in the number and the nature of the sample genotypes and testing environment or a combination of both might explain these contradictory results on the relative importance of N uptake and N utilization.

After examining the feasibility of breeding for N efficiency in the early 1990s (Lafitte and Edmeades, 1994), CIMMYT developed N-efficiency screening protocols (Bänziger et al., 2000) and incorporated breeding for N efficiency in its routine breeding programs (Beck et al., 1996). Selection is conducted simultaneously under unstressed conditions and a range of managed abiotic and biotic stress conditions, including low-N, representing highest priority stresses in the target environment of a particular breeding program. This approach has been shown to lead to significant selection gains in random stress environments in sub-Saharan Africa, including low-N environments (Bänziger et al., 2004). So far, no information on the underlying physiological mechanisms is available for the hybrids and open-pollinated varieties resulting from this breeding approach.

This study assesses a set of contrasting N-efficient tropical maize hybrids, adapted to the midaltitudes in sub-Saharan Africa, across a range of N levels. The study examines to what extent N uptake and N utilization contribute to N efficiency in this set of hybrids and the relevance of G x E for grain yield formation at different N levels. Establishing distinct N efficiency parameters of individual hybrids will assist in further examining the underlying physiological mechanisms for N efficiency in future studies.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Hybrids
Fourteen single-cross and three-way cross maize hybrids bred at CIMMYT stations in Kenya, Mexico, and Zimbabwe and two commercial maize hybrids (one single-cross and one three-way cross hybrid) from the breeding program of Seed Co International in Zimbabwe were used, including three Quality Protein Maize (QPM) hybrids (Table 1). Even though originating from different breeding programs, in previous experiments, all hybrids had shown adaptation to the midaltitude areas (900–1700 m above sea level, Corbett, 1998) of eastern and southern Africa. On the basis of data collected within CIMMYT's breeding program, the grain yield of these hybrids differed under low-N, indicating potential differences in N efficiency characteristics.


View this table:
[in this window]
[in a new window]

  		Table 1. Maize hybrids used in this study.

 
Experimental Sites and Trial Management
The experiments were conducted at the CIMMYT research station at Harare, Zimbabwe, (2003 and 2004) and at the Kenya Agricultural Research Institute research station at Kiboko (2003); these represented subhumid and dry midaltitude areas of sub-Saharan Africa (Corbett, 1998; Hornetz et al., 2000). The research site at Harare is located at 31°1‘ S, 17°49’ E and 1478 m above sea level. The soil is a reddish brown clay soil (nitosol) with a bulk density of 1.4, 1.5, and 1.5 g cm–3 for 0- to 30-cm, 30- to 60-cm and 60- to 90-cm depth, respectively. The pH (CaCl2) of the topsoil (0–30 cm) and subsoil (30–60 cm) was 5.89 and 5.93, respectively. The annual monomodal rainfall averaged 684 mm (2003 and 2004) and maximum and minimum temperature during the growing season (October to April 2004) averaged 28.3 and 14.2°C, respectively. The site at Kiboko, Kenya, is located at 2°10‘ S, 37°40’ E at an elevation of 975 m above sea level. Long-term annual rainfall is 561 mm in a bimodal distribution (328 mm from October to January; 233 mm from March to June). Average, maximum and minimum temperatures are 33 and 14°C from March to August, respectively. The soil is a eutric fluvisol with sandy-clay texture. The pH (CaCl2) of the topsoil was 7.9, decreasing to 5.8 in the subsoil (70–80 cm). The high pH of the topsoil was attributed to the Kiboko river irrigation water (Hornetz et al., 2000). The bulk density of the soil was 1.5, 1.6, and 1.6 g cm–3 for 0- to 30-cm, 30- to 60-cm and 60- to 90-cm depth, respectively. The amounts of nitrogen available in the form of nitrate at planting are presented in Table 2.


View this table:
[in this window]
[in a new window]

  		Table 2. Fertilizer application (N1, N2, N3), cropping history, and soil nitrate N content at planting of the testing environments in Harare, Zimbabwe, 2003 and 2004 (Z03, Z04) and Kiboko, Kenya, 2003 (K03).

 
The hybrids were tested under three N levels (low-N, medium-N, and high-N) at both sites, resulting in nine environments: Harare 2003 low-N (Z03N1), Harare 2003 medium-N (Z03N2), Harare 2003 high-N (Z03N3), Kiboko low-N (K03N1), Kiboko medium-N (K03N2), Kiboko high-N (K03N3), Harare 2004 low-N (Z04N1), Harare 2004 medium-N (Z04N2), and Harare 2004 high-N (Z04N3). Given different soil types, cropping histories, and environmental conditions at each station, it did not seem feasible to establish the exact same N supply conditions at each site. Instead choice of N application and previous cropping history was intended to create a range of relevant soil-N levels at each station, ranging from severe N stress to N levels which were not limited by N (Table 2). Depletion under low N was achieved by continuously growing maize for a minimum of 5 yr and applying no (Harare) or <20 kg ha–1 of N (Kiboko). High N rates were split and applied at planting and 40 d after planting. Phosphorus and potassium were applied uniformly on the basis of site-specific recommendations before planting.

Trials at both sites were hand-planted with two seeds per hill and thinned at the 3-leaf stage. Planting dates at Harare were on 3, 7, and 4 Dec. 2002 for Z03N1, Z03N2, and Z03N3 and on 26, 28, and 27 Nov. 2003 for Z04N1, Z04N2, and Z04N3, respectively. Harvest was between 31 April 2003 and 5 May 2003 in 2003 and between 16 April 2004 and 25 April 2004 in 2004. The trials at Kiboko were planted on 14 April 2003 and harvested between 27 Aug. 2003 and 1 Sep. 2003. Pre-emergence herbicides—atrazine (6-chloro-N'-ethyl-N-propan-2-yl-1,3,5-triazine-2,4-diamine, 4.5 L ha–1) and metalchlor (96%) [2-choloro-6'-ethyl-N-(2-methoxy-1-methylethyl)acet-o-toluidide 1.8 L ha–1]—were applied at planting to control the weeds. Then, the weeds were controlled by hand weeding and application of Basagran (BASF, Corp., 3 L ha–1; 48% bentazon: 3-isopropyl(1H)-benso-2,1,3-tiadiazin-4-on-2,2-dioxid). Furadan (FMC Corp., 20 kg ha–1, carbofuran: 2,3-dihydro-2,2-dimethylbenzofuran-7-yl methylcarbamate) was applied at planting. Fungal diseases [Cercospora zeae-maydis Tehon & E.Y. Daniels, Exserohilum turcicum (Pass.) K.J. Leonard & E.G. Suggs, and Puccinia sorghi Schwein.] were controlled by Tilt 250EC (Syngenta, 0.5 L ha–1). Thiodan 1G (Hoechst Schering AgrEvo Pty Ltd, 4 kg ha–1) and Thionex (Makhteshim-Agan Pty Ltd, 230 g ha–1, endosulfan) were used to control stalk borers [Busseola fusca (Fuller) and Chilo partellus (Swinhoe)]. Cutworms were controlled with lambda-cyhalothrin {(R)-{propto}-cyano-3-phenoxybenzyl (1S)-cis-3-[(Z)-2-chloro-3,3,3-trifluoropropenyl]-2,2-dimethylcyclopropanecarboxylate and (S)-{propto} -cyano-3-phenoxybenzyl (1R)-cis-3-[(Z)-2-chloro-3,3,3-trifluoropropenyl]-2,2-dimethylcyclopropanecarboxylate), 5 g ha–1} applied at emergence. The trials at Harare were irrigated to field capacity at planting by sprinkler irrigation. A second irrigation of 20 to 30 mm was applied 6 to 7 d after planting to facilitate germination. Thereafter, trials were irrigated to field capacity whenever soil moisture was less than 40% of field capacity. Similar procedures were followed for trials at Kiboko. A plot size of 4-m length by 4.5-m width with six rows per plot was used. Spacing was 0.75 and 0.25 m between rows and plants, respectively. A plant density of 53333 plants per hectare was kept after thinning.

Measurements
Anthesis date (AD) was recorded for each plot when 50% of the plants in the two central rows shed pollen. On the following day, 12 plants were harvested from an area of 2.25 m2 in the central four rows for dry matter determination and plant N analysis at anthesis. An area of 5.65 m2, corresponding to 32 plants in the central four rows, was harvested immediately after physiological maturity for dry matter determination and plant N analysis at harvest. Both harvest areas were bordered by a minimum of two plants per row on each side of the harvested areas; this has been shown to be sufficient to exclude significant border experiments in previous experiments (Bänziger et al., 2000).

Grain yield was recorded from all ears in the harvest area at physiological maturity. Ears were shelled, grain weight and grain moisture percentage were recorded, and grain yield (Mg ha–1) calculated at 125 g kg–1 moisture. Fresh weight of the plants harvested at anthesis and physiological maturity was recorded after cutting plants at ground level and, in the case of the harvest at physiological maturity, removing the ears. One quarter of the plants together with a quarter of the shelled cobs (for the harvest at physiological maturity) were finely chopped with a VIKING 220 chopping machine (VIKING GmbH, Langkampfen, Austria), producing a homogenized stover mix of which a subsample of 1 to 2 kg was taken. Stover subsample and grain sample for each plot were weighed, oven-dried to constant weight at 80°C for 72 h, weighed again, and total stover, grain, and plant biomass was calculated. Grain and stover subsamples were milled with an analytical mill, sent to Hanover, and analyzed for N at the Plant Nutrition Institute Laboratory, University of Hanover, Germany, with a CNS analyzer (Vario EL, Elementar Analysis systems, Hanau, Germany).

Nitrogen uptake, utilization and harvest index were calculated from data taken at physiological maturity. Nitrogen uptake (NUP) was set equal to the total N (kg) in the aboveground biomass. Nitrogen utilization (NUT) was calculated as the ratio of dry grain yield (kg) to total N (kg) in the above ground biomass. Nitrogen harvest index (NHI) was calculated as the ratio of N found in the dry grain to total N in the aboveground biomass expressed as a percentage. In addition, N uptake at anthesis (NUA), N uptake between anthesis and physiological maturity (NUAP), and stover N at physiological maturity (STONP) were calculated. Plant height (height from ground level to the base of the tassel), and stem circumference (at 6 cm above the ground) were measured.

Experimental Design and Statistical Analysis
The hybrids were planted under alpha (0,1) lattice designs (Patterson and Williams, 1976) with four replications. Within each experiment, lattice-adjusted hybrid means were calculated by the PROC MIXED procedure of SAS (SAS Institute, 2001), with hybrids as a fixed factor and replicate and incomplete blocks within replicates as random factors.

For the G x E interaction analysis, it was first established whether G x E was at all significant in this set of experiments with hybrids as fixed factors and experiments as random factors, i.e., the significance for genotype mean square was tested against the G x E interaction mean square, while the significance of G x E mean square was tested against the pooled error. Genotype x environment interactions were then further analyzed by additive main effect and multiplicative interaction (AMMI) analysis (Crossa et al., 1990; Gauch, 1992; Ebdon and Gauch, 2002) to assess similarity and dissimilarity among testing environments and interaction patterns between hybrids and environments. Biplots of the first two interaction principal component analysis axes (IPCA1 and IPCA2) were used to illustrate the patterns. Close points within environments and hybrids indicate similarity while distant points indicate dissimilarity. Hybrids and environments that are close together also tend to have similar interaction patterns (Gauch, 1992; Fox et al., 1997; Betran et al., 2003). AMMI analysis was computed by SAS (SAS Institute, 2001). Simple linear regression coefficients were also calculated for each hybrid by regressing the yield of individual hybrid on mean grain yield of the experiment. Slopes of the regression were tested for significant differences from unity by t tests (Eberhart and Russell, 1966). Finally, simple linear correlation coefficients were calculated to assess the association between traits in each environment.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Nitrogen Stress Intensity and Grain Yield
The nine environments significantly (P < 0.01) varied in grain yield, N uptake, N utilization, and N harvest index (Tables 3 and 4). Z03N1 was the lowest yielding environment, while Z03N3 was the highest yielding environment. Severe stress under low-N at Harare reduced grain yield by 65 and 77% in 2003 and by 47 and 70% in 2004 as compared with medium-N and high-N treatments, respectively. At Kiboko, low-N stress in K03N1 reduced grain yield by 25 and 40% as compared with K03N2 and K03N3, respectively, indicating that the severity of low-N stress was less than at Harare. Grain yields of hybrids ranged from 1.5 to 4.3, 6.1 to 9.8, and 10.6 to 14.9 Mg ha–1 in 2003 and 2.2 to 4.4, 4.6 to 9.4, and 8.7 to 13.5 Mg ha–1 in 2004 for Harare low-N, medium-N, and high-N experiments, respectively, while they ranged from 3.7 to 7.3, 5.8 to 10.9, and 5.5 to 13.2 Mg ha–1 for Kiboko low-N, medium-N, and high-N experiments, respectively. Differences between the hybrids and hybrid x environment interactions were significant (P < 0.01) (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 3. Mean grain yield of each testing environment.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Hybrid means, statistics, and factor effects for anthesis date (AD); N uptake in the aboveground biomass at anthesis (NUA); and physiological maturity (NUP), N utilization (NUT), N harvest index (NHI), and grain yield (GY). In addition, slope (b) of the linear regression for hybrid grain yield on the mean grain yield of each experiment is presented.

 
Genotype x Environment Interactions
AMMI biplots for grain yield (Fig. 1 ) showed distinct patterns for the testing environments. Axis 1 and 2 of the Interaction Principal Component Analysis (IPCA1 and IPCA2) were significant (P < 0.01) and explained 37.3 and 28.6%, respectively, of the G x E interaction. Together the two principal components explained 65.9% of the G x E variation. The AMMI analysis separated Harare and Kiboko environments; similar N levels were little (Z03N1 and Z04N1 versus K03N1; Z03N3 and Z04N3 versus K03N3) or negatively (Z03N2 and Z04N2 versus K03N2) correlated, indicating different environmental effects at Harare and Kiboko for these experiments. At each location, the angles between N1, N2, and N3 showed less relationship between N1 and N3 as compared with N1 and N2 or N2 and N3. Indeed, good performance under low-N showed little to no correlation with good performance under high-N at each location.


Figure 1
View larger version (15K):
[in this window]
[in a new window]

  		Figure 1. Additive main effect and multiplicative interaction (AMMI) biplots for grain yield of 16 hybrids (H) evaluated under three N levels each at Harare, Zimbabwe, 2003 (Z03N1, Z03N2, Z03N3), Kiboko, Kenya, 2003 (K03N1, K03N2, K03N3) and Harare, Zimbabwe, 2004 (Z04N1, Z04N2, Z04N3). IPCA, interaction principal component analysis axis.

 
Projection of hybrid points on environmental vectors made it possible to distinguish between hybrids mostly favored in low-N and those mostly favored in high-N environments. Hybrids 4, 9, 11, and 16 projected positively on all three low-N vectors and, expect for Hybrid 11, projected negatively on all three high-N vectors, indicating relatively better performance under low-N and relatively poorer performance under high-N. Hybrids 5, 7, and 14 projected positively on all three high-N vectors and negatively on all three low-N vectors and, thus, showed relatively better performance under low-N and relatively poorer performance under high-N. This was also confirmed by the b-value for the linear regression of grain yield of individual hybrids on mean grain yield of each experiment (Table 4). Hybrids 4, 9, 11, and 16 were among those with a low b-value, and Hybrids 5, 7, and 14 were among those with a high b-value. Hybrids 1 and 10 and to a minor extent Hybrids 2 and 3 showed specific adaptation to all N levels at Harare (Fig. 1). Hybrids 8, 11, 12, and 13 showed specific adaptation to all N levels at Kiboko.

AMMI analysis expresses relative better or poorer performance of a hybrid as compared with its overall mean. It does not indicate which hybrid was the best performing in a particular environment. Focusing on the trials in Zimbabwe where mean grain yields between low-N and high-N levels differed most and AMMI biplots showed a stronger dissimilarity of high and low N environments (Table 3 and Fig. 1), hybrid performance under low-N was plotted against hybrid performance under high-N (Fig. 2 ), which made it possible to distinguish between efficient and inefficient hybrids on the basis of above-average and below-average performance under low-N, respectively, and responsive and nonresponsive hybrids on the basis of above-average and below-average performance under high-N, respectively (Sattelmacher et al., 1994). Hybrid 7 was among the highest yielding hybrids at Harare high-N (Z03N3 and Z04N3), while it was among the inefficient hybrids at Harare low-N (Z03N1 and Z04N1). In comparison, Hybrids 5, 6, and 15 had comparable grain yields to Hybrid 7 under high-N conditions but relatively higher yields under low-N; i.e., they were more efficient (Fig. 2). Hybrids 16 and 12 were the most efficient hybrids, but they were less responsive than with Hybrid 7.


Figure 2
View larger version (19K):
[in this window]
[in a new window]

  		Figure 2. Relationships between grain yields of 16 maize hybrids (H) at Harare high-N and low-N. Results are means of Harare 2003 and 2004. Broken lines represent mean grain yield.

 
Nitrogen Uptake and Utilization Efficiencies
Nitrogen in the aboveground biomass increased as N level in the soil increased. At maturity, it varied from 47 kg N ha–1 at Z03N1 to 278 kg N ha–1 at Z03N3 (Fig. 3 ). It was higher at Harare high-N in both years as compared with Kiboko high-N, but N uptake was higher at Kiboko low-N as compared with Harare low-N in both years. More N was accumulated at Harare medium-N in 2003 than at Harare medium-N in 2004 and Kiboko medium-N in 2003. Significant differences (P < 0.01) were observed among the hybrids for N uptake in all environments (data not shown).


Figure 3
View larger version (29K):
[in this window]
[in a new window]

  		Figure 3. Mean N uptake (NUP) of 16 hybrids evaluated under three N levels each at Harare, Zimbabwe, 2003 (Z03N1, Z03N2, Z03N3), Kiboko, Kenya, 2003 (K03N1, K03N2, K03N3) and Harare, Zimbabwe, 2004 (Z04N1, Z04N2, Z04N3).

 
Under low-N conditions, the correlation between grain yield and N uptake at physiological maturity was always positive and close at both sites (Table 5). A significantly positive relationship was observed between N uptake between anthesis and physiological maturity (NUAP) and grain yield under low-N conditions, whereas the correlation between N uptake at anthesis (NUA) and grain yield was insignificant.


View this table:
[in this window]
[in a new window]

  		Table 5.  Linear correlation coefficients (r) between grain yield (GY) and N uptake at anthesis (NUA), N uptake after anthesis (NUAP), stover N at physiological maturity (STONP), total N uptake in the aboveground biomass at physiological maturity (NUP), N utilization (NUT), N harvest index (NHI), and anthesis date (AD) in the testing environments differing in levels of N stress.

 
At Harare, the efficient and responsive Hybrid 15 was among the hybrids with a high N uptake under low-N conditions (Fig. 4A ). The responsive hybrid with poor performance under low-N, Hybrid 7, had a high N uptake only under high-N condition indicating crossover interaction for N uptake as N stress increased. Efficient Hybrids 12 and 16 expressed high-N uptake under low-N conditions and below-average N-uptake under high-N conditions. Under low-N stress of Harare (Z03N1 and Z04N1), the difference between total N uptake at maturity and total N uptake at anthesis was positive for most of the efficient hybrids, while it was negative for most of the inefficient hybrids (data not shown).


Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		Figure 4. Relationships between (A.) N uptake (NUP), (B.) N utilization (NUT, kg dry grain/kg N uptake), and (C.) N harvest index (NHI) of contrasting N-efficient maize hybrids (H) at Harare high-N and Harare low-N. Results are means of 2003 and 2004. Broken lines represent mean values for each trait.

 
Although hybrids varied significantly (P < 0.01) for days to anthesis, anthesis date was significantly related to grain yield only at Z03N2 and Z04N2 (Table 5). Similarly, the relation between N uptake and anthesis date, plant height and stem circumference was low under low-N stress (data not shown) implying that these traits were less important for N-uptake in this group of hybrids and environments. However, there was a strong relationship between N-uptake and total above-ground biomass at physiological maturity with correlation coefficients (r) ranging from 0.78 (significant at P < 0.01) to 0.92 (significant at P < 0.01) in different environments (data not shown).

Significant differences (P < 0.01) were observed among the hybrids for N utilization and N harvest index (Table 4). Nitrogen utilization was also significantly positively related with grain yield under severe low-N stress (Table 5) implying that the efficient hybrids generally produce more grain per unit N taken up under severe low-N stress as compared with the inefficient hybrids (Fig. 4B). The efficient and less responsive hybrids, like Hybrid 1 and Hybrid 2 had high N utilization under low-N conditions but less N uptake under high-N condition as compared with the responsive hybrids (Fig. 4A). Among the efficient and responsive hybrids, Hybrid 12 had a high N utilization under both low-N and high-N (Fig. 4A). The inefficient and responsive Hybrid 7 had high N utilization under high-N condition only, indicating the crossover interaction for N utilization under different environments. Generally, most of the hybrids with medium N utilization had good N uptake (Fig. 4A, B). The relation between grain yield and N utilization was positive in all environments except at K03N2. Mean N utilization (NUT) progressively declined as the mean N uptake in the total aboveground biomass increased (data not shown).

Nitrogen harvest index was positively related to grain yield in all environments (Table 5). Most of the hybrids with high N utilization, for example Hybrid 16 and Hybrid 12, also showed a high N harvest index under low-N conditions (Fig. 4C). The responsive inefficient Hybrid 7 achieved the highest N harvest index under high-N conditions and an average N harvest index under low-N conditions. The inefficient hybrid, Hybrid 13, was among the hybrids with a medium and high N uptake under low-N and high-N, respectively (Fig. 4A), but it was among the hybrids with low N utilization (Fig. 4B) and N harvest index (Fig. 4C), indicating the importance of N utilization and N partitioning for grain yield performance under all conditions.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Studies with large number of genotypes established that genetic correlation between maize grain yields under low-N and high-N is generally positive but decreases with increasing relative yield reduction under low-N (Bänziger et al., 1997; Presterl et al., 2003). In the present study, experiments were conducted at two locations (Harare and Kiboko) and three N levels (low-N, medium-N, and high-N) at each. Even though AMMI analysis showed hybrid performance to be somewhat different between the two locations, this study confirmed a different discrimination of hybrid grain yield under low-N than under high-N at both Kiboko and Harare (Fig. 1). Grain yield under medium-N and high-N conditions and medium-N and low-N conditions were each related, and grain yields of low-N and high-N experiments were somewhat more closely related at Kiboko where low-N stress was relatively less than at Harare (yield reduction of 40% at Kiboko versus 70–77% at Harare). Hence, hybrids in this study followed the general pattern of N stress exhibiting an increasing G x E interaction as the difference in N stress intensity between two environments increased (Bänziger et al., 1997; Presterl et al., 2003). These results may have been affected by the fact that the hybrids in this study were selected for contrasting performance under low N; however, Bänziger et al. (1997) and Presterl et al. (2003) both established their results in unselected populations of random progenies.

Given that low-N is one of the most limiting factors for maize production in sub-Saharan Africa where many farmers grow maize with little to no N fertilizer application on heavily depleted soils (Sibale and Smith, 1997; Bänziger et al., 2004), the experiments with the most severe N stress were of greatest interest in this study. Graham (1984) and Sattelmacher et al. (1994) defined as N efficiency the ability of a genotype to realize an above-average grain yield under conditions of low-N availability. In the present study, both N uptake and N utilization contributed to N efficiency. This is in contrast to the results of Moll et al. (1982) who reported that N utilization is more important than N uptake under low-N conditions and Kamprath et al. (1982) who reported N uptake is more important than N utilization under low-N conditions. Presterl et al. (2002) reported positive relationships among grain yield and N uptake and N utilization under low-N conditions, but these authors mainly related N efficiency to improved N uptake under low-N. The results of this study are in agreement with the results of Wiesler et al. (2001) for temperate maize and Ortiz-Monasterio et al. (2001) for wheat; these authors found that both N utilization and N uptake were important traits for improved performance under low-N conditions.

The relationship between N uptake and N efficiency was mostly related to differences in postanthesis N uptake. Except for one experiment (medium-N at Kiboko 2003), there was no relation between N uptake before anthesis and grain yield, whereas N uptake after anthesis was consistently positively related to grain yield under severe low-N conditions (Table 5). Similar results were found in temperate maize, oilseed rape (Brassica napus L., Wiesler et al., 2001), and in tropical maize (Akintoye et al., 1999) where the authors concluded that maintaining root activity for continued N absorption during grain filling was important for N efficiency.

Partitioning of N within the plant and efficient utilization of N at the cellular level are contributing factors for N utilization under limited N condition (Sattelmacher et al., 1994). Hence, even though preventing N remobilization from the vegetative parts of the plant is desirable for continued photosynthesis (Bänziger et al., 2002), appropriate N partitioning between grain and stover is needed for grain yield formation under N stress. Ta and Weiland (1992) found genotypic variation in translocation of stored N from the vegetative part to the grain sink. In our study, N harvest index was correlated with N efficiency to a similar extent as were N uptake and N utilization, indicating that N efficient hybrids remobilized relatively more N from vegetative plant parts than did inefficient hybrids.

Causes of variation in N efficiency in terms of component factors might differ between levels of N supply and among genotypes (Moll et al., 1982; Ortiz-Monasterio et al., 2001), as was also reflected in this study. Different from low-N conditions, however, the relative contribution of increased N uptake or N utilization to increased grain yield under medium-N or high-N was not consistent between locations and years. N uptake and N utilization were uniquely important in one experiment each—N uptake at medium-N in Kiboko 2003 and N utilization at medium-N in Harare 2003—whereas, in all other experiments, both factors were related to the higher grain yield of hybrids. In all experiments where N uptake was significant, N uptake between anthesis and physiological maturity was more important than N uptake before anthesis for increased grain yields.

Among the hybrids evaluated, AMMI analysis identified several hybrids with specific adaptation to either low-N or high-N conditions, as compared with their average performance across experiments. Several hybrids exhibited specific adaptation to either Kiboko or Harare, which implied that the location of evaluation had to be considered when comparing hybrids. At Harare, Hybrid 7 was among the highest yielding hybrids under high-N conditions, with high N uptake, N utilization, and N partitioning. When grown under low-N, the yield of Hybrid 7 dropped relative to several other hybrids that were associated by a strong decrease in N uptake. Hybrids 2, 9, 12, 15, and 16 were high yielding under low-N conditions at Harare. This was related under low-N to the high N uptake for Hybrid 15, high N utilization for Hybrid 2, and high N uptake and N utilization for Hybrids 9, 12, and 16.

The study included single-cross and double-cross hybrids (Table 1), and there was also some variation in anthesis date among hybrids (Table 4). No consistent effect of these factors on the results of this study could be discriminated.

In conclusion, post-flowering N uptake and N utilization contributed to improved performance under low-N conditions in this set of tropical maize hybrids, whereas N uptake before anthesis was of little relevance. The study confirmed the relevance of G x E interactions for grain yield formation in maize as N stress intensity increases but also identified some hybrids with high yields under both low-N and high-N conditions. More detailed studies are required to determine the underlying physiological mechanisms contributing to the N efficiency in selected hybrids.

Received for publication May 15, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
H. G. Gauch Jr., H.-P. Piepho, and P. Annicchiarico
Statistical Analysis of Yield Trials by AMMI and GGE: Further Considerations
Crop Sci., May 1, 2008; 48(3): 866 - 889.
[Abstract] [Full Text] [PDF]

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 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Agricola
Right arrow Articles by Worku, M.
Right arrow Articles by Horst, W. J.
Related Collections
Right arrow Nitrogen
Right arrow Maize
Right arrow Crop Genetics

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