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


     

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 HighWire
Right arrow Citing Articles via ISI Web of Science (26)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Agricola
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Related Collections
Right arrow Crop Genetics
Right arrow Plant and Environment Interactions
Crop Science 43:807-817 (2003)
© 2003 Crop Science Society of America

CROP BREEDING, GENETICS & CYTOLOGY

Genetic Analysis of Inbred and Hybrid Grain Yield under Stress and Nonstress Environments in Tropical Maize

F. J. Betrán*,a, D. Beckb, M. Bänzigerc and G. O. Edmeadesd

a Corn Breeding and Genetics Program, Texas A&M University, College Station, TX77845
b International Maize and Wheat Improvement Center (CIMMYT), Apdo. Postal 6-641, 06600 México D.F., México
c International Maize and Wheat Improvement Center (CIMMYT). P.O. Box MP163, Harare, Zimbabwe
d Pioneer Hi-Bred Int., Box 609, Waimea, HI 96796

* Corresponding author (javier-betran{at}tamu.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Drought and low soil N cause significant yield reductions in maize (Zea mays L.) grown in the tropics. Understanding the genetic basis of hybrid performance under these stresses is crucial to designing appropriate breeding strategies. This study evaluates under optimal, drought and low N stress conditions (i) the performance, combining abilities and stability of a group of tropical white inbred lines; (ii) the genetic control and modes of gene action for grain yield; and (iii) the relationship between line per se and hybrid performance. Seventeen lowland white-grained tropical maize inbred lines were used in a diallel study. Lines and their hybrids were evaluated separately in trials under drought stress, low N, and optimal conditions in a total of 12 environments. The differences in grain yield between hybrids and inbreds (i.e., heterosis) increased with the intensity of drought stress. Significant interactions were observed for combining abilities under low and high N. The type of gene action appeared to be different under drought than under low N, with additive effects more important under drought and dominance effects more important under low N. The importance of additive effects increased with intensity of drought stress. This suggests the need for drought tolerance in both parental lines to achieve acceptable hybrid performance under severe drought. Inbreds derived from the population ‘La Posta Sequía’ exhibited the highest GCA effects, stability coefficients, and frequency of dominant alleles for grain yield. Good performance across stress levels can be achieved in tropical maize hybrids.

Abbreviations: AMMI, Additive Main Effects and Multiplicative Interaction • ASI, anthesis-silking interval • CML, CIMMYT maize line • DT, drought tolerance • GCA, general combining ability • IS, intermediate stress • LPS, ‘La Posta Sequía’ • QTL, quantitative trait loci • SCA, specific combining ability • SS, Severe stress • TS, ‘Tuxpeño Sequía’ • WW, well watered


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE PRODUCTION OF ADEQUATE FOOD in the tropics is being challenged by rapid population growth (80 million people/year) and declining availability of water resources and arable and fertile land (Beck et al., 1997). Drought and low N stresses are factors most frequently limiting maize production in the tropics. Maize grain losses due to drought in the tropics may average as much as 24 million megagrams per year, equivalent to 17% of well-watered production (Edmeades et al., 1992). Possible climate change due to global warming could further increase the chances of drought. Low N availability in soils is an important yield-limiting factor frequently found in farmers' fields in the tropics where fertilization is not commonly used and organic matter is rapidly mineralized (Bänziger and Lafitte, 1997). The development of maize germplasm able to tolerate drought and low N stress is crucial if the productivity of maize-based farming systems is to be sustained or increased.

Maize population improvement for drought tolerance (DT) at flowering has been accomplished in source populations by recurrent selection using managed drought stress (Bolaños and Edmeades, 1993a; Edmeades et al., 1999). Recurrent selection under drought was effective at increasing yield across a range of drought stress levels in all populations under evaluation (Edmeades et al., 1999). Gains under stressed conditions were significantly greater than those observed in conventionally selected counterpart populations without loss of yield potential (Byrne et al., 1995). Improvements were due to a significant reduction in barrenness and increases in grain number per ear and harvest index, and were accompanied by a reduction in the anthesis-silking interval (ASI) and a delay in leaf senescence (Bolaños and Edmeades, 1993b; Edmeades et al., 1999). Improvement for drought tolerance also brought specific adaptation and improved performance under low N conditions suggesting that tolerance to either stress involves common adaptive mechanisms (Bänziger et al., 1999).

Edmeades et al. (1997) showed that population improvement for stress tolerance in source populations increases the probability of deriving stress-tolerant hybrids from those populations. CIMMYT has used these populations as source germplasm to develop inbred lines and hybrids (Beck et al., 1996). Breeding strategies to develop stress tolerant maize inbred lines include screening and selection of inbreds under managed stress conditions, multilocation testing of progenies in a representative sample of the target environments, and selection under high plant populations (Beck et al., 1996). Additional information from adaptive secondary traits that show differential expression between optimal and stress conditions are genetically variable and are correlated with grain yield (e.g., barrenness, ASI, foliar senescence; Bolaños et al., 1993; Bolaños and Edmeades, 1996; Bänziger and Lafitte, 1997) is commonly used to increase the selection efficiency.

In the transition from developing stress tolerant source populations to developing stress tolerant hybrids, important issues that need to be considered include the relationship between inbred and hybrid performance under stress, the comparative performance of inbreds selected for stress tolerance versus conventionally selected inbreds, gene action and dosage effects, and the type of tester most appropriate for developing stress tolerant hybrids. The present study evaluates the following under drought, low N, and well-watered, well-fertilized conditions: (i) the performance, combining abilities, and stability of a group of tropical white inbred lines selected conventionally or selected under managed drought stress; (ii) the genetic control and modes of gene action for grain yield; and (iii) the relationship between line per se and hybrid performance.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Germplasm
Seventeen white-grained lowland tropical maize inbred lines were used in a diallel study (Table 1) . These lines were developed according to different selection criteria. Nine inbreds were developed from populations improved for drought tolerance by recurrent selection in controlled drought stress environments [La Posta Sequía C3 (LP) and Tuxpeño Sequía 6 C1 (TS)] (Edmeades et al., 1999). Some of these inbreds were released as CIMMYT Maize Lines (CMLs) at the conclusion of this study (Table 1). They were developed by pedigree breeding with evaluation of topcross performance under drought stressed and well-watered conditions. A selection index for standardized variables across environments was used to select the best lines in hybrid combination across water regimes (Beck et al., 1997). The remaining seven inbreds were elite CMLs selected for combining ability and agronomic performance across several nonstress environments (Table 1). During this study, CML247 and CML 254, which produce a stable high yielding hybrid, were the principal tropical white testers used in the CIMMYT program. Lastly, the experimental line P1 was selected for short ASI under drought, and was used to map quantitative trait loci (QTL) for ASI (Ribaut et al., 1996; 1997). Diallel crosses among the lines were made in the 1995 summer and winter seasons at the CIMMYT experiment station in Poza Rica, México. Seeds from reciprocal crosses of the full diallel were bulked to form a set of 136 F1 hybrids.


View this table:
[in this window]
[in a new window]
  		Table 1. Maize inbred lines involved in the study, their pedigrees, and the primary selection criteria used in their development.

 
Evaluation and Stress Management
Lines and hybrids were evaluated separately in two trials planted side by side in 12 environments (Table 2) . Experimental designs were {alpha} (0,1) lattice designs (Patterson and Williams, 1976) with incomplete block sizes of six plots for the lines and 17 plots for the hybrids. Hybrids and lines were oversown with two seeds per hill every 20 cm (33 cm in OB96BISHS and 40 cm in TL97AISLD) in single rows of 2.5 or 5 m in length spaced 75 cm apart, which were later thinned to the desired plant densities (Table 2). All trials received standard cultural practices to control insects and weeds.


View this table:
[in this window]
[in a new window]
  		Table 2. Characteristics of the experimental evaluations conducted for inbred lines and their hybrids.

 
Six environments (TL96AWW, PR96AWW, CO96AWW, TL96BWW, PR96BHN, CO97BWW) comprised optimal fertilization and supplemental irrigation as needed to avoid moisture stress. The trials were conducted at Tlaltizapán, Morelos, México (18°41' N, 940 m above sea level), Poza Rica, Veracruz, México (21°N, 60 m above sea level) and Cotaxtla, Veracruz, México (18°50' N, 60 m above sea level) during the winter and summer seasons. Five experiments (TL96ASS, TL96AIS, OB96BISHS, TL97AISHD, TL97AISLD) were conducted under partial drought stress at Tlaltizapán during the winter season and at Cd. Obregón in the Sonoran Desert, México (27°30' N, 30 m above sea level) during the summer season. Both locations are largely rain free during these seasons, allowing the control of drought stress intensity by withdrawing or delaying irrigation for varying lengths of time during flowering and grain filling stages (Edmeades et al., 1999). Intermediate drought stress was achieved by withholding water from 2 wk before silking to the end of the flowering period. Severe drought stress (TL96ASS) was achieved by withholding water from 4 to 5 wk before silking to the end of the flowering period. Poza Rica was also used to grow an experiment (PR96BLN) at reduced natural N levels in the field achieved by continuous cropping of maize without N fertilizer application. A short-stature sorghum [Sorghum bicolor (L.) Moench] was intercropped between the maize rows and was cut and removed 2 wk before flowering of the maize crop to increase N deficiency (Bänziger et al., 1997). Apart from the targeted stress, the management of trials at the same locations was the same across irrigation and N levels.

Field Measurements
Plot ear weight was measured in the field with a manual scale at CO96AWW, OB96BISHS, and CO97BWW. A shelling percentage of 80% was assumed when grain yields were determined. Grain moisture was measured electronically from representative grain samples. In all other trials, ears were dried to constant grain moisture and shelled. All grain weights were adjusted to 155 g kg-1 moisture content. In trials under stress (TL96ASS, TL96AIS, OB96BISHS, PR96BLN, TL97AISHD, TL97AISLD), where border effects were obvious, one border plant next to the alley was discarded.

Statistical Analysis
Individual analyses of variance were conducted for each trial with the PROC MIXED procedure from SAS (SAS, 1997) with genotypes (hybrids or inbreds) being considered as fixed effects, and reps and blocks within reps random effects. If the incomplete block design was more efficient than a randomized complete block design, the adjusted means were used to estimate general combining ability (GCA) and specific combining ability (SCA) effects.

Repeatabilities were calculated as the proportion of genetic variance over the total phenotypic variance (Fehr, 1987). They represent an upper limit for broad-sense heritabilities. Griffing's method IV diallel analysis was used to estimate GCA for the lines and SCA for the hybrids in individual environments and across environments (Griffing, 1956). GCA and SCA equivalent variance components of mean squares were calculated by a fixed model for the diallel design (Baker, 1978). The relative importance of general and specific combining ability on progeny performance was estimated as the ratio:

where, 2GCA, 2SCA are the variance components for GCA and SCA.

Average degree of dominance and parental order of dominance across environments was estimated by means of the Vr-Wr graphical analysis of Hayman (1954). Vr is the variance of all the progenies in each parental array. Wr is the covariance between parents and their offspring in each array. When Wr is regressed on Vr, the degree of dominance can be characterized by the sign (+/-) and magnitude of the Y-intercept: a = 0 means complete dominance, a > 0 means partial dominance, and a < 0 means overdominance. The parental order of dominance can be estimated by the (Wr +Vr) values. High values indicate low frequency of dominant alleles and vice versa. Likewise parents close to the origin contain a higher frequency of dominant alleles.

The degree of correlation between inbred line per se and hybrid performance was estimated by regressing the F1 hybrid values on midparent values. Simple phenotypic correlation coefficients among test environments for the same trait were computed. Combined analyses of variance were conducted by means of PROC GLM in SAS (SAS, 1997). Because error variances for grain yield were not homogeneous across all environments according to Bartlett's test (P = 0.05), the environments were grouped by a hierarchical agglomerative clustering with Ward's minimum variance as a method of fusion. Joint linear regression was used to estimate stability of inbred lines (Eberhart and Russell, 1966). Grain yield for inbred lines per se and in hybrid combination was used in the analysis. Additive Main Effects and Multiplicative Interaction (AMMI) analysis was computed using the same mean values to assess relations among inbreds, environments, and between inbreds and environments (Crossa, 1990; Gauch and Zobel, 1988). A SAS program from the CIMMYT Biometrics Unit was used in the AMMI computations (Vargas and Crossa, 2000). Biplots of the first two principal components were used to illustrate these relationships (Gabriel, 1971; Kempton, 1984). Environments are represented as vectors. Inbreds and environments that are close together tend to be similar. The angle between two environments indicates the degree of association or correlation. Small angles indicate similarity, 90° angles indicate orthogonality and no association, and angles >90° indicate a negative correlation of genotype performance between these environments. The orthogonal projections of inbred genotypes on environment vectors indicate the relative performance of inbreds in a given environment; that is, the greater the projection of the genotype in the positive direction, the better the performance of that inbred in that environment.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Hybrid and Inbred Grain Yields
Genotype and genotype x environment (GxE) interactions were significant for grain yield of hybrids and inbred lines. Mean squares for genotypes, GxE interaction, and pooled error were 22.82, 2.55, and 0.72, respectively. Mean grain yields for hybrids ranged from 1.14 Mg ha-1 in TLSS under severe drought stress to 9.18 Mg ha-1 in TLSWW under subtropical and well-watered conditions during the summer season (Table 2). Grain yields for inbred lines ranged from 0.15 Mg ha-1 in TL96ASS under severe drought stress to 3.95 Mg ha-1 in PR96AWW under tropical and well-watered conditions during the winter season. The average grain yield across environments was 6.01 and 2.27 Mg ha-1 for hybrids and inbreds, respectively. At Tlaltizapán, grain yields of hybrids tested under severe (TL96ASS) and intermediate (TL96AIS) drought stress during the 1996 winter season were 13 and 50% of grain yield under well-watered conditions (TL96AWW), respectively. Grain yield of the lines under severe (TL96ASS) and intermediate (TL96AIS) drought stress during the same season were 5 and 48% of their grain yield under well-watered (TLWW) conditions, respectively. At Poza Rica grain yield of hybrids under low N (PR96BLN) during the 1996 summer season averaged 33% of that under high N (PR96BHN). Grain yield for inbred lines under low N (PR96BLN) was 65% of that under high N (PR96BHN). Such levels of intensity of stress observed for drought and low soil nitrogen fall within the range of stress levels applied during selection of populations and inbred lines for tolerance to drought or low N (Bolaños and Edmeades, 1993a; Lafitte and Edmeades, 1994). Compared with hybrids, inbred lines were relatively more sensitive to drought stress than low N stress. Maturity differences were detected both among the 17 late tropical inbreds and hybrids. Phenotypic correlations between grain yield and male and female flowering dates were mostly negative in both inbreds and hybrids across both stress and nonstress environments (data not shown).

The highest yielding hybrids across locations were TS4 x LP2, CML258 x LP5, and CML258 x LP1; under drought stress CML254 x LP2, CML258 x LP5, and CML258 x LP1; and under low N TS4 x LP2, CML254 x LP5, and CML254 x LP5. Inbred lines with different selection history produced high yielding hybrids under contrasting environments, indicating that inbreds selected for drought stress tolerance can provide sources with different frequencies of alleles that combine effectively with alleles present in conventionally selected inbreds. Furthermore, some hybrids performed well across stress levels, indicating that it is possible to combine stress tolerance and yield potential in tropical maize hybrids. Similar results have been reported with temperate maize hybrids where improvements for tolerance to abiotic and biotic stresses have been associated with the ability to maximize grain yield under nonstress growing conditions (Carlone and Russell, 1987; Castleberry et al., 1984; Duvick, 1997).

General Combining Ability of the Lines
GCA and GCA x environment interaction effects were significant for grain yield. Mean squares for GCA and GCA x environment interaction were 82.91 and 13.09, respectively. LP2 had the highest GCA effects under drought and across environments (Table 3) . Inbreds developed from La Posta Sequía (LPS) had greater GCA effects and grain yield than did inbreds developed from Tuxpeno Sequía (TS) and most of the conventionally selected inbreds. This relative superior performance of LPS inbreds over TS inbreds was consistent across environments, except for COWI and PRLN (Fig. 1) . The selection method applied during the development of the drought tolerant lines and their source populations was similar and consisted of a selection index that included grain yield in stressed and nonstressed environments, as well as secondary traits (Beck et al., 1997). LPS inbreds were developed after three cycles of S1 recurrent selection while TS inbreds were developed after six cycles of full-sib plus one cycle of S1 recurrent selection. TS has a longer breeding history for drought tolerance than LPS but the performance of LPS inbreds per se and in hybrid combinations was better than that of TS inbreds. The GCA effects of conventionally selected inbreds were less consistent across environments than those of TS and LPS inbreds. CML 258 had the second highest GCA across environments as well as under intermediate drought stress. CML 254 showed consistent positive GCA effects across most environments while CML 247 showed negative GCA effects. Tall highly prolific inbred LP1 had the highest grain yield across environments as inbred line per se followed by CML 274, LP5, and LP4. SCA was significant, while its interaction with environments (SCAxE) was not. Mean squares for SCA and SCAxE were 14.74 and 1.12, respectively. SCA was negative for crosses involving inbreds with the same genetic background, while it was positive for crosses involving inbreds with different genetic background (data not shown) (Betrán et al., 2003).


View this table:
[in this window]
[in a new window]
  		Table 3. General combining ability effects (GCA) and inbred line per se performance (Inbred) for grain yield (GY) (Mg ha-1) by environment and across environments, and correlation between GCA and inbred per se for grain yield GY.

 


View larger version (37K):
[in this window]
[in a new window]
  		Fig. 1. General combining ability effect estimates (GCA) for grain yield across 12 environments for 17 tropical white maize inbred lines. Environments are 1: TL96AS, 2: TL96AIS, 3: TL96AWW, 4: PR96AWW, 5: CO96AWW, 6: OB96BISHS, 7: TL96BWW, 8: PR96BHN, 9: PR96BLN, 10: CO96BWW, 11: TL97AISHD, and 12: TL97AISLD.

 
Gene Action and Hayman Analysis
Repeatabilities were quite consistent across environments and stress levels (data not shown). The GCA and SCA genetic variance components for grain yield were smaller for stressed environments than for well-watered environments. The relative importance of GCA vs. SCA, expressed as the ratio between additive vs. total genetic variance components, increased with drought stress level when comparing trials grown at the same location and during the same season (Fig. 2) . Additive variance accounted for 84% of the genetic variance under severe drought stress (TL96ASS) and 60% under well-watered conditions (TL96AWW). This difference is significant on the basis of the standard errors associated with the additive variance component. Additive genetic effects across environments, which accounted for 61% of total genetic variation, seem to be more important than nonadditive genetic effects. These ratios of additive to nonadditive variances are similar to the average reported in temperate maize (Hallauer and Miranda, 1988). Under low N, nonadditive genetic effects were more important than the additive genetic effects and a significant number of crossovers were observed between the GCA of lines under low N and high N. Results in past evaluations of tropical maize progenies also indicate that when low N stress became more severe the correlation of grain yields between low and high N decreased and changes in genotype ranks increased when compared with high N environments (Bänziger et al., 1997). The gene action modulating grain yield under drought in this set of hybrids seems to be different from that under low N, with additive effects more important under drought. This suggests the need for drought tolerance in both parental lines to obtain acceptable hybrid performance under severe drought stress that coincides with flowering. The importance of finding inbreds with high SCA becomes more critical under low N.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Proportion of additive (lower bar) and nonadditive (upper bar) genetic variance for grain yield at 12 stress and nonstress environments and across environments in a diallel among 17 tropical maize inbreds.

 
In the Hayman graphical analysis, the Y-intercept for the Wr-Vr regression was –0.412, indicating the presence of overdominance (Fig. 3) . This is a reflection of the difference in performance between hybrids and inbreds across environments. The intercepts per environment ranged from –0.959 (TL97AISHD) to 0.227 (PR96BLN). The ranking of the inbreds from high to low frequency of dominant alleles for grain yield was LP1 > LP3 > CML254 > P1 > LP2 > CML273 > CML274 > CML247 > CML258 > CML268 > LP5 > TS1 > CML264 > LP4 > TS4 > TS2 > TS5. LPS-derived inbreds provided a greater frequency of dominant alleles for grain yield compared with TS-derived inbreds. The superior performance of LPS inbreds with high frequency of dominant alleles make them good candidates parents for recycling inbred lines.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 3. Covariance between parental inbreds and hybrids (Wr) vs. variance of all the hybrids in each parental array (Vr) in Hayman's graph for grain yield across environments in a diallel among 17 tropical maize inbreds (CMLs, TS inbreds, LP inbreds).

 
Correlation between Inbred Line Per Se and Hybrid Performance
The possibility of using inbred line information to indicate hybrid performance under stress could reduce the need for hybrid evaluation. The correlation between hybrid and midparent grain yield was positive and significant in all the locations, except in PR96BHN and TL97AISLD (Fig. 4) . A lower correlation was observed under severe drought stress than under optimal conditions. Contrasting results were obtained in an earlier study when examining the relationship between 100 S3 lines derived from TS6 C1 and 100 from LPS C3 and their topcrosses with Tuxpeño and non-Tuxpeño testers under SS, IS, and WW (Betrán et al., 1995). Traits of inbred lines under SS in that study were more strongly correlated with topcross performance under severe drought stress than under non-stress normal conditions. In the present study the poor performance of inbreds under severe drought stress (0.15 Mg ha-1) could contribute to the weak correlation. The degree of inbreeding of the lines (e.g., S3 vs. homozygous inbreds) could also contribute to the weak correlation, since high drought intensity can be tolerated by early generation inbred lines (e.g., S2 and S3) than by homozygous inbreds. A different trend was observed for low N where the correlation between line and hybrid performance was greater (0.33) than under high N (0.08). The correlation under high N (PR96BHN), however, was unusually low compared with other nonstress environments (TL96AWW, PR96AWW, CO96AWW, TL96BWW, CO96BWW). This was inconsistent with the type of predominant gene action observed in the hybrids, since high correlations should be indicative of additive gene action. However, the gene action was estimated on the basis of hybrid performance alone while the correlation incorporates the inbred line per se performance, which depends on the degree of stress applied. El-Lakany and Russell (1971) reported that the inbred–hybrid correlation was greater under stress due to increasing plant densities but has not been important under low N conditions (Lafitte and Edmeades, 1995; Balko and Russell, 1980). Stronger correlations have been reported for other traits such as maturity, plant height, and 100-kernel weight under low N stress (Lafitte and Edmeades, 1995). Correlation between parent and offspring has often been reported to increase in extreme environments as the average environmental contribution to the phenotype increases (Ward, 1994). The correlations observed in this study were similar in magnitude to those reported in the literature for grain yield (Hallauer and Miranda, 1988). Despite the positive correlations between inbreds and hybrids in all the environments, comparative yield trials of the hybrids are still needed, especially in light of the variation in correlations across environments.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 4. Midparent (inbreds)-offspring (hybrid) correlation for grain yield (Mg/ha) at 12 stress and nonstress environments and across environments in a diallel among 17 tropical maize inbreds.

 
Relationship among Environments
Both the genetic correlations among environments (data not shown) and the cluster analysis for grain yield of hybrids group the environments similarly (Fig. 5) . The groups were mainly defined by megaenvironments separating summer and winter and subtropical and tropical environments. In winter environments, it takes longer to reach maturity because of a slower rate of accumulation of growing degree units. Also, the drought stress environments were conducted in the winter, which corresponds to the dry season. The environments with different stress level but sharing the same location were clustered together indicating a similar ranking of this set of genotypes across stress levels. The exception was PRHN and PRLN that were clustered in different groups.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 5. Cluster analysis (Ward's Minimum Variance) of 12 stress and nonstress environments based on grain yield of hybrids in a diallel among 17 tropical maize inbreds.

 
In growing areas in the tropics where drought stress is common, farmers reduce the application of N fertilizer. Bänziger et al. (1999) reported that selection for drought stress increased the grain yield under N stress in lowland tropical maize populations. The identification of hybrids with superior performance in both stresses would enhance maize production in the tropics. The drought stress and low N environments were clustered in different groups in our study. This could indicate different response of hybrids to these two stresses and the need of specific development of hybrids for each stress. However, drought stress environment were conducted in subtropical environments during the winter while low N environments were conducted in tropical environments during the summer. Therefore, we cannot conclude that the different response of hybrids to drought and Low N is due to different adaptive mechanisms.

The variation in other environmental variables, such as photoperiod, climatic, and edaphic conditions, seemed to have stronger influence in this set of genotypes than differential stress levels applied in a single managed environment. Genetic correlations for grain yield measured in stress and nonstress environments have been sufficiently low in other studies to make direct selection under stress more effective in maize than indirect selection under optimal conditions (Brun and Dudley, 1989; Byrne et al., 1995; Bänziger et al., 1997) and other cereals (Atlin and Frey, 1990; Ceccarelli et al., 1992; Ud-Din et al., 1992; Zavala-Garcia et al., 1992). The germplasm under selection or evaluation plays a critical role in defining relationships among environments and ultimately in the relative efficiency of direct vs. indirect selection. Some drought tolerant inbred lines in this study also had the highest GCA under optimal conditions (Table 3) and this may have influenced the clustering of environments in this study. We conclude that selection would be more efficient if stress environments were combined with testing in target locations so GxE interaction can be estimated.

GCA and Line Per Se AMMI and Stability Analysis
AMMI's biplots for average grain yield in hybrids (e.g., GCA effects) and for inbred per se grain yields (Fig. 6) positioned the inbreds according to the locations where they were primarily selected. The reaction across environments was similar for inbreds from the same breeding group in both cases. AMMI principal component 1 was significant and explained 34.3 and 35% of the GxE variation in hybrids and inbreds. AMMI principal component 2 was also significant and explained 23.9 and 29.5% of the variation in hybrids and inbreds. LP4 was the inbred with the most stable GCA effects across environments. Conventionally selected inbreds performed better in tropical environments (Poza Rica and Cotaxtla), where they were selected. LPS and TS inbreds showed more affinity for subtropical locations (Tlaltizapan and Obregon). They were developed by screening for drought tolerance at these sites. The response of conventionally selected inbreds was more diverse than that of LPS and TS inbreds that grouped together more consistently.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 6. Additive Main Effects and Multiplicative Interaction (AMMI) biplots for grain yield in hybrids (GCA effects) and inbred lines in a diallel among 17 tropical maize inbreds ({circ} CMLs, {blacktriangleup} TS, LP) evaluated in 12 stress and non-stress environments ({diamondsuit}).

 
Stability regression coefficients (Eberhart and Russell, 1966) for average grain yield in hybrids and inbred lines were in general >1 for inbreds showing the best combining ability and per se performance as inbreds (LPS derived inbreds, CML254, CML258, CML268), and <1 for inbreds with low values for GCA and per se performance (TS and remaining CMLs) (Table 4) . Inbreds with high mean grain yield showed regression coefficients greater than one and greater deviation from linearity. Association of high mean yield and regression coefficients have been found in temperate maize (Eberhart and Russell, 1966; Gama and Hallauer, 1980). The stability parameters for GCA indicate the average ability of the inbreds to respond to the environment across hybrid combinations. Inbreds with high GCA and regression coefficients >1 should indicate good response to favorable environments, corresponding to the "agronomic concept of stability" (Becker, 1981). High yielding inbreds increased both hybrid and inbred per se grain yield at a greater rate than observed in poor inbreds when environmental conditions improved. This suggests that it is possible to develop good inbreds for both stress and optimal conditions, and confirms the value of the selection strategy used to develop these drought tolerant inbreds, i.e., to improve drought tolerance while maintaining the capability to perform under optimal conditions.


View this table:
[in this window]
[in a new window]
  		Table 4. Average grain yield (Mg ha-1), phenotypic stability, and deviation from linearity for 17 tropical inbreds per se and in hybrid combination.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Tropical white inbred lines were evaluated for per se performance and in hybrid combinations across a set of stress and nonstress subtropical and tropical environments. The genetic parameters estimated varied with the location and degree of stress. Differences in grain yield between hybrids and inbreds increased with the intensity of drought stress. Compared with hybrids, inbred lines were relatively more sensitive to drought than to low N stress. Additive effects were more important under drought than under low N stress, and increased in importance as the severity of drought increased. This suggests potential benefits of incorporating drought tolerance in both parental inbreds to enhance hybrid performance under drought. Inbreds derived from LPS imparted a higher frequency of dominant alleles for grain yield. This together with their superior performance, GCA effects, and stability coefficients make these inbreds good candidates for direct use as parents in hybrids and new breeding populations (e.g., recycling). The correlation between F1 and midparent grain yield was significant but variable across environments.

Genetic variance and heritability of grain yield and therefore expected breeding progress during the development of maize inbreds are generally lower under stress than under optimal conditions (Bänziger and Cooper, 2001). However, the genetic correlation for grain yield between stress and optimal environments seems to decrease as stress intensity increases (Bänziger et al., 1997; Cooper et al., 1997; Fukai et al., 1999). This suggests that selection in optimal environments could not be effective in identifying superior genotypes for stress environments. Predicted selection under managed stress conditions appears to be more efficient than indirect selection under nonstress conditions when targeting stress environments (Bänziger and Cooper, 2001). In this study, tropical white inbreds selected for drought tolerance performed well under stress and nonstressed conditions and they combined well with other inbreds selected for stability and optimal conditions, providing evidence that good performance across stress levels can be achieved for tropical maize hybrids.


   ACKNOWLEDGMENTS
 
We thank Srs. Ezequiel Bahena, Oscar Fernández, Fernando González, and their assistants who conducted the field trials in a dedicated and careful manner. Thanks to Ing. Ciro Sánchez for his assistance in the execution of the lab and field activities. We appreciate the assistance and guidance by Drs. José Crossa, Jorge Franco and Mateo Vargas in data analysis. Thanks to all the personnel at the CIMMYT research stations in Tlaltizapán and Poza Rica, and to the INIFAP research stations in Obregón and Cotaxtla. This research was partly funded by the United Nations Development Program (UNDP).

Received for publication December 20, 2001.


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




This article has been cited by other articles:


Home page
 Crop Sci.Home page
M. Worku, M. Banziger, G. S. a. Erley, D. Friesen, A. O. Diallo, and W. J. Horst
Nitrogen Uptake and Utilization in Contrasting Nitrogen Efficient Tropical Maize Hybrids
Crop Sci., March 1, 2007; 47(2): 519 - 528.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
H. G. Gauch Jr.
Statistical Analysis of Yield Trials by AMMI and GGE
Crop Sci., May 18, 2006; 46(4): 1488 - 1500.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
K. E. D'Andrea, M. E. Otegui, A. G. Cirilo, and G. Eyherabide
Genotypic Variability in Morphological and Physiological Traits among Maize Inbred Lines--Nitrogen Responses
Crop Sci., April 25, 2006; 46(3): 1266 - 1276.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
C. Welcker, C. The, B. Andreau, C. De Leon, S. N. Parentoni, J. Bernal, J. Felicite, C. Zonkeng, F. Salazar, L. Narro, et al.
Heterosis and Combining Ability for Maize Adaptation to Tropical Acid Soils: Implications for Future Breeding Strategies
Crop Sci., October 27, 2005; 45(6): 2405 - 2413.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
F. J. Betran, J. M. Ribaut, D. Beck, and D. G. de Leon
Genetic Diversity, Specific Combining Ability, and Heterosis in Tropical Maize under Stress and Nonstress Environments
Crop Sci., May 1, 2003; 43(3): 797 - 806.
[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 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 HighWire
Right arrow Citing Articles via ISI Web of Science (26)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Agricola
Right arrow Articles by Betrán, F. J.
Right arrow Articles by Edmeades, G. O.
Related Collections
Right arrow Crop Genetics
Right arrow Plant and Environment Interactions

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Plant Registrations Soil Science Society of America Journal
Journal of Natural Resources
and Life Sciences Education		 Journal of
Environmental Quality