Published in Crop Sci. 44:741-747 (2004).
© 2004 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP BREEDING, GENETICS & CYTOLOGY
Yield Evaluation of Maize Cultivars across Environments with Different Levels of Pink Stem Borer Infestation
A. Butrón*,
P. Velasco,
A. Ordás and
R. A. Malvar
Misión Biológica de Galicia, Spanish Council for Scientific Research, Apartado 28, 36080 Pontevedra, Spain
* Corresponding author (abutron{at}mbg.cesga.es).
 |
ABSTRACT
|
|---|
Yield under infestation by the pink stem borer [Sesamia nonagrioides (Lefebvre)] has been proposed as the best estimator of maize (Zea mays L.) performance under pink stem borer attack. Yield is a complex trait that greatly interacts with the environment. Several methods could be used to study the genotype x environment (GE) interaction. The objective of this work was to study the GE interaction for yield of 49 maize hybrids in five different environments by the Site Regression (SREG) and factorial regression methods. Locations presented different levels of natural infestation by the pink stem borer. The biplot obtained by applying the SREG method allowed visual cultivar evaluation. The factorial regression method incorporated genotypic and environmental covariates that enhanced biological interpretation of GE interaction. Hybrid A637 x EP42 would be recommended in northwestern Spain under medium and high natural infestation by the pink stem borer because it showed high and stable yielding ability. Genotypic and environmental covariates explained approximately 75% of the GE interaction variation, but other genotypic covariates could be introduced in the model to reduce GE residual variation to a nonsignificant level. In general, GE effects for grain yield were mainly due to earliness, vigor effects, and environmental yield limiting factors such as the mean of minimum temperature and percentage of air humidity.
Abbreviations: E, Environmental main effects • G, Genotype main effects • GE, Genotype x environment • GGE, G plus GE interaction • SREG, Sites Regression Model
|
INTRODUCTION
|
|---|
THE MOST IMPORTANT PEST of maize in the temperate areas of the northern hemisphere is the European corn borer [Ostrinia nubilalis (Hübner)] (Dicke and Guthrie, 1988), but, in the Mediterranean area, the pink stem borer also causes significant damage to maize (Anglade, 1972). Specifically, in the northwest of Spain, the pink stem borer is the main pest of maize (Cordero et al., 1998). Two environmental factors determine the pink stem borer development, temperature and feed (Hilal, 1981). The pink stem borer is a tropical moth (Fam. Noctuidae) and its population levels are limited at temperatures below 0°C (Galichet, 1982).
The larvae bore tunnels in maize stalks, destroying the pith and weakening the plant, and, consequently, reducing grain yield (Anglade, 1961; García-Martí et al., 1996). Larue (1984) reported that yield losses by pink stem borer could reach up to 30%. Maize yield losses in northwestern Spain averaged on a set of 45 hybrids were 15% of the insect free crop (Butrón et al., 1999). As the relationship between maize resistance and yield loss was very weak, it was thought that it would be convenient to select genotypes by a comprehensive measure such as yield loss that depends on tolerance and antibiosis of genotypes (Butrón et al., 1998). Nevertheless, some genotypes that suffer important yield losses could achieve high yield under infestation because of their high yield potential (Lynch, 1980; Butrón et al., 1999). Yield is a complex trait that greatly interacts with the environment. Although pink stem borer infestation could play a role in the interaction, other environmental factors should be considered.
Several methods could be used to study the GE interaction (Crossa, 1990; Brancourt-Hulmel et al., 1997). Methods, which include multiplicative and factorial regression models, appear as the most suitable for interpretation of the response of genotypes to different environments. Those approaches describe the interaction multiplicatively as a genotype score multiplied by an environment score. The difference between the multiplicative models, such as the Additive Main Effects and Multiplicative Interaction (AMMI) and Sites Regression (SREG) (Crossa and Cornelius, 1997), and the factorial regression is that in the multiplicative model both parameters are unknown, while only a single parameter is unknown in regression (Brancourt-Hulmel et al., 1997; Vargas et al., 1998). Factorial regression incorporates genotypic and environmental covariates (Denis, 1988) that enhance biological interpretation of genotype (G) and environmental (E) main effects, and GE interaction, but it requires relevant covariates (Baril et al., 1995; Brancourt-Hulmel et al., 1997). Multiplicative models elucidates a different part of the interaction than factorial regression, and both methods have been considered as complementary (Baril, 1992; Van Eeuwijk et al., 1995).
Among multiplicative models, the SREG method (Crossa and Cornelius, 1997) has been suggested as the appropriate model for analyzing multienvironment trials when large yield variation is due to environments (Yan et al., 2000). The SREG method supplies a graphical display called GGE (G plus GE interaction) biplot that facilitates visual cultivar evaluation. The objective of this work was to study the GE interaction of forty-five maize hybrids tested in five different environments characterized by different levels of natural infestation by the pink stem borer.
|
MATERIALS AND METHODS
|
|---|
Ten inbreds with different degrees of antibiosis to stem tunneling by the pink stem borer (Table 1) were crossed to produce a diallel without reciprocals. The 45 F1 single crosses were evaluated along with four checks in a 7 x 7 triple lattice design under natural infestation by pink stem borer. Experiments were performed in five different environments: in 1995 at two places, Pontevedra (42°24' N, 8°38' W, 20 m above sea level) and Pontecaldelas (42° 23' N, 8° 32' W, 300 m above sea level), and in 1996 at three locations, Pontevedra, Pontecaldelas, and Ribadumia (42° 30' N, 8° 46' W, 50 m above sea level). Locations were in northwestern Spain and presented different levels of natural infestation (0.54, 0.91, 1.78, 2.62, and 3.35 larvae per plant were found at Pontecaldelas, 1995, Pontecaldelas, 1996, Ribadumia, 1996, Pontevedra, 1995, and Pontevedra, 1996, respectively, when about 750 plants were dissected in each environment). Pink stem borer damage was only important at Pontevedra and Ribadumia.
Each two-row experimental plot consisted of 13 hills with two kernels per hill. The rows were spaced 0.80 m apart and the hills were spaced 0.21 m apart. Hills were thinned to one plant after emergence, obtaining a final plant density of 60 000 plants ha–1.
Data from each plot were recorded on early vigor when plants were in the five-leaf stage by giving visual ratings that ranged from 1 (weak) to 5 (vigorous), days to pollen shed (days from planting until 50% of plants shed pollen), and days to silking (days from planting until 50% of plants show silks). At harvest, five stems per plot were dissected and tunnel length (cm) was recorded. The general appearance of five ears per plot was rated on a five point scale from 1 (ears without damage) to 5 (wholly damaged ears). Yield (Mg/ha–1) as shelled grain weight at 140 g kg–1 moisture per plot was recorded. Vigor, days to pollen shed, days to silking, tunnel length, and general appearance of the ear averaged across environments were selected as genotype covariates.
An environment was defined as the combination of a year and a location. Locations were not more than 20 km apart from each other, but differed for other climatic parameters besides pink stem borer infestation. Parameters related to temperature, rainfall, and air humidity were recorded in each environment to obtain a better biological explanation of interaction effects. Environmental covariates considered were average number of larvae per plant across genotypes, average daily temperature, mean of daily minimum temperatures, mean of daily maximum temperatures, minimum temperature, maximum temperature, percentage of air humidity, and rainfall.
The fixed effect two-way model for analyzing multienvironments cultivar trials is as follow:
where 
i and ßj are the genotype and environmental main effects, (ß)ij is the GE interaction effect, and 
ij is the residual experimental error.
The sites regression model used was (Cornelius et al., 1996; Crossa and Cornelius, 1997):
where n goes from 1 to r, with r = number of principal components (PCs) required to approximate the original data. 
*in and 
*jn are the ith genotype and the jth environmental scores for PCn, respectively. In the SREG method, PC analysis is made on residuals of an additive model with environments as the only main effects. Therefore, the term 
*in *jn contains the variation due to G and GE interactions. A two-dimensional biplot (Gabriel, 1971) called GGE biplot (G plus GE interaction) of the two first PCs was plotted (Yan et al., 2000). Genotypes and environments were displayed in the same plot. Each genotype and environment was defined by the genotype's and environment's scores on the two PCs, respectively. Analysis were made by a SAS (SAS, 2000) program for graphing GE and GGE biplots developed by Burgueño et al. (2003).
Factorial Regression
The general form for a factorial regression model with K genotypic and H environmental covariates is (Denis, 1980):
where 
k and 
h are the regression coefficients of genotypic Gik, and environmental covariates Ejk, respectively; i and ßj are the residuals of genotype and environmental main effects, respectively; 
kh is the regression coefficient of the cross-product of covariates Gik and Ejh; and 'ih and ß'jk are the genotype i and environment j specific regression coefficient of environmental covariate Ejh and genotypic covariate Gik, respectively. The residual interaction effect is ij. All parameters were considered fixed. The covariates and their order in the factor regression model for yield data were obtained by performing a stepwise regression independently on genotype and environmental covariates, without considering the other set of covariates (Denis, 1988). After standardization of covariates, factorial regression analyses were performed by the computer package INTERA (Decoux and Denis, 1991). All terms were tested against the residual experimental error, which resulted from the combination of individual errors of triple 7 x 7 lattice designs (Cochran and Cox, 1957).
|
RESULTS AND DISCUSSION
|
|---|
The analysis of variance showed that the environment main effect (E) was the most important source of variation, accounting for 50% of total sum of squares for yield (Table 2). About 38% of variation was explained by GGE. Genotype main effects (G) accounted for the 66% of the GGE variation (Table 3). Therefore, variation due to G was larger than that due to the GE interaction, but GE interaction was significant, meaning that differences among genotypes vary across environments. Van Eeuwijk et al. (1995) also found that variation due to the GE interaction was small in relation to the G variation for silage dry matter content of 18 Dutch maize varieties. Epinat-Le Signor et al. (2001), however, reported higher GE interaction than G for grain yield in a study with early maize hybrids tested in about 30 locations in northern France. The first five PCs obtained after singular value decomposition of location-centered yield data were significant and explained 100% of GGE variation (Table 2). The first two PCs of the SREG model explained 78.62% of GGE variation. Means for grain yield for each genotype and each environment as well as the two first PCs appear in Table 4.
View this table:
[in this window]
[in a new window]
|
Table 2. Analysis of variance of the SREG multiplicative model for yield of 49 hybrids evaluated in five different environments with different levels of natural infestation by the pink stem borer.
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Means for grain yield in each environment and averaged across environments, and PC1 and PC2 scores from the SREG analysis.
|
|
Yan et al. (2000) and Crossa et al. (2002) stated that, in the two-dimensional biplot (Fig. 1)
, if the primary effects of sites from the SREG model are all of the same sign as it was in the present study (Table 4), PC1 presents a noncrossover GE interaction. A genotype with a larger PC1 score has a greater average yield and its performance varies across environments in direct proportion to the environment PC1 scores. The two-dimensional biplot showed that the hybrid A661 x EP42 had the highest yield at Pontecaldelas and Pontevedra in 1996 and at Pontecaldelas in 1995, although A637 x EP42 performed as well as A661 x EP42 in those environments (Fig. 1) (Table 4). Besides, the hybrid A637 x EP42 was the most productive in Ribadumia 1996 and Pontevedra 1995 as well. The hybrid CM105 x PB60 also presented high yield in the five environments. All three hybrids cited above were more productive than the check hybrids and followed the heterotic pattern American dent x European flint extensively used in Europe (Moreno-González, 1988). On the contrary, crosses to inbred lines A509, EP28, EP31, F7, and Z77016 had, in general, negative PC1 scores, suggesting poor average performance. The hybrid F7 x Z77016 had the worst PC1 score. F7 and Z77016 could be related in origin because the cross F7 x Z77016 yielded as little as inbred lines (data not shown). Therefore, hybrid A637 x EP42 could be recommended to maize growers for its high yield under medium and high natural infestation by the pink stem borer, although either A637 or EP42 presented unfavorable GCA effects for tunnel length in a previous work (Butrón et al., 1999).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. The GGE biplot based on maize hybrid yield performance under pink stem borer infestation in five environments. Hybrids of the diallel were designated by the letters assigned to their inbred parents: A, B, C, D, E, F, G, H, I, and J for A509, A637, A661, CM105, EP28, EP31, EP42, F7, PB60, and Z77016, respectively. The hybrid checks were designated T.
|
|
The PC2 represents disproportionate genotype yield differences across environments. The genotype A661 x EP42 had moderate negative interactions with Pontevedra 1995 and Ribadumia 1996 as the PC2 scores for this hybrid and the locations were of opposite sign, and positive interactions with the remaining environments since PC2 scores were of the same sign (Fig. 1). On the other hand, the low genotypic PC2 score for A637 x EP42 represents proportionate response of the genotype across environments. Therefore, A637 x EP42 appeared as a high- and stable-yielding genotype because it showed a large PC1 score and a near-zero PC2 score. On the contrary, F7 x Z77016 was stable, but presented low yield in all environments.
Factorial regression was performed to obtain a biological explanation of the G, E, and GE interaction. The number of larvae per plant, mean of daily minimum temperatures, and percentage of air humidity were the only environmental covariates that were detected as significant by the stepwise method (data not shown). On the other hand, days to silking, early vigor, and tunnel length were retained among the genotypic covariates.
The three significant environmental covariates explained almost the total variation for E (Table 3). The number of larvae per plant and the mean of daily minimum temperature were equally important in explaining E. The regression coefficients of the environmental main effects on environmental covariates (n) showed that grain yield increased with increased number of larvae per plant, and with decreased mean daily minimum temperature and average air humidity (Table 3). Although pink stem borer attack is a limiting factor for obtaining the potential maize yield, the environmental conditions that were favorable for maize growth were also favorable for larvae development and could, therefore, compensate yield losses by insect attack. The direct relationship between mean daily minimum temperature and yield was positive (data not shown), but became negative after removing the linear effect of the averaged number of larvae per plant on the environmental main effect variation for yield (Table 3). Besides pink stem borer attack, there were other stress factors such as weeds. The incidence of weeds was much more important at Ribadumia and Pontevedra than at Pontecaldelas and could explain the negative relationship between yield and mean minimum daily temperature; Pontecaldelas was cooler than the other two locations.
The proportion of the G explained by genotypic covariates was only 45.7%. The variation for days to silking explained 31% of yield variability among genotypes, while each of the other two genotypic covariates explained less than 10% of genotypic variability. The regression coefficients of G on genotypic covariates (n) showed positive relationships of grain yield with days to silking, and early vigor. Nevertheless, increased tunnel length was associated with higher yield (Table 3). This finding corroborated the lack of a strong relationship between resistance to the borer and yield losses (Butrón et al., 1998).
The factorial regression model explained almost 75% of the GE interaction sum of squares for maize grain yield, although the residual GE was significant (Table 3). Other authors have reported on factorial regression models which accounted for a high proportion of the interaction variation, without reducing residual interaction variation to nonsignificant levels (Hébert et al., 1995; Biarnès-Dumoulin et al., 1996; Epinat-Le Signor et al., 2001). The nine genotypic x environmental covariate cross-products explained a small part of sum of squares for the interaction (3.6%) and only two covariate cross-products were significant (Table 3). The negative regression parameter of yield on early vigor x mean daily minimum temperature (V-TM) showed that the interaction effect of vigorous varieties increased with lower mean daily minimum temperatures, while the interaction effect of less vigorous hybrids increased as mean daily minimum temperature increased (Table 3). That was expected since maize development is lower with temperatures below the optimal threshold for maize growth, but that is less obvious for genotypes with higher vigor at early stages because they are better adapted to cool temperatures (Revilla et al., 2000). The regression coefficient of yield on the cross-product days to silking x air humidity (S-H) was negative, indicating that the interaction effect for early genotypes increased as air humidity increased, while late genotypes were favored under lower air humidity (Table 3). This could be interpreted as the advantage of early genotypes over late genotypes to loose grain humidity when the percentage of air humidity is high.
Although two covariate cross-products were significant, most of the variability for the interaction was explained by the interaction of environmental covariates with the residual genotype variation (65%). Only a small proportion of the sum of squares for the interaction was explained by the interaction of genotype covariates with the residual environmental variation (3.5%). There were significant genotype-specific responses to environmental covariates, which could not be explained by differences in any of the genotype covariates used. Yield also showed significant differences among environment-specific responses to days to silking and early vigor, but not to tunnel length (Table 3). Therefore, there was no significant yield reduction due to increased tunnel length in any environment.
GE effects for grain yield were mainly due to earliness, vigor effects and environmental yield limiting factors such as the mean of minimum temperature and percentage of air humidity. Other authors have reported that GE interaction effects for silage and grain yield were mainly due to earliness effects and yield limiting factors (Argillier et al., 1994; Epinat-Le Signor et al., 2001).
Genotypic resistance to pink stem borer attack did not seem to have any influence on GE interaction for grain yield and did not improve genotypic performance. Therefore, selection for resistance to pink stem borer attack may not result in yield improvement. Other authors have even reported losses of yielding ability after selection for insect resistance (Russell et al., 1979; Klenke et al., 1988; Nyhus et al., 1989; Butrón et al., 2002).
In conclusion, hybrid A637 x EP42 would be recommended in northwestern Spain under medium and high natural infestation by the pink stem borer because it showed high and stable yielding ability. Genotypic and environmental covariates explained, approximately, 75% of the GE interaction variation, but other genotypic covariates could be introduced in the model to reduce GE residual variation to a nonsignificant level. In general, GE effects for grain yield were mainly due to earliness, vigor effects and environmental yield limiting factors such as the mean of minimum temperature and percentage of air humidity.
|
ACKNOWLEDGMENTS
|
|---|
We thank Dr. Moro for his invaluable help in the statistical processing of data and for his careful review and Dr. Denis for facilitating the access to the statistical package INTERA.
|
NOTES
|
|---|
Research supported by the Plan Nacional I+D+I (AGL2000-0944-C02-01). A. Butrón acknowledges a fellowship from the High Council for Scientific Research that allowed her to carry out this study.
Received for publication March 18, 2003.
|
REFERENCES
|
|---|
- Anglade, P. 1961. Influence sur le rendement du maïs de l'infestation des tiges par la deuxième génération de la Sésamie (Sesamia nonagrioides Lef. Lep. Noctuidae). Méthodes de comparaison des hybrides par infestation artificielle. Ann. Epiph. 12:357–372.
- Anglade, P. 1972. Les sesamia. p. 1389–1400. In A.S. Balachowsky (ed.) Entomologie appliquée à l'agriculture. II, Lépidoptères, II. Masson et Cie, Paris.
- Argillier, O., Y. Hébert, and Y. Barrière. 1994. Statistical analysis and interpretation of line x environment interaction for biomass yield in maize. Agronomie 14:661–672.
- Baril, C.P. 1992. Factor regression for interpreting genotype-environment interaction in bread-wheat trials. Theor. Appl. Genet. 83:1022–1026.
- Baril, C.P., J.-B. Denis, R. Wustman, and F.A. van Eeuwijk. 1995. Analysing genotype by environment interaction in Dutch potato variety trials using factorial regression. Euphytica 82:149–155.
- Biarnès-Dumoulin, V., J.-B. Denis, I. Lejeune-Hénaut, and G. Etévé. 1996. Interpreting yield instability in pea using genotypic and environmental covariates. Crop Sci. 36:115–120.[Abstract/Free Full Text]
- Brancourt-Hulmel, M., V. Biarnès-Dumoulin, and J.B. Denis. 1997. Points de repère dans l'analyse de la stabilité et de l'interaction génotype-milieu en amélioration des plantes. Agronomie 17:219–246.
- Burgueño, J., J. Crossa, and M. Vargas. 2003. SAS programs for graphing GE and GGE biplots. Downloaded from http://www.Cimmyt.org/Research/Biometrics/bsu.htm; verified 15 January 2004.
- Butrón, A., R.A. Malvar, P. Velasco, P. Revilla, and A. Ordás. 1998. Defense mechanisms of maize against pink stem borer. Crop Sci. 38:1159–1163.[Abstract/Free Full Text]
- Butrón, A., R.A. Malvar, P. Velasco, M.I. Vales, and A. Ordás. 1999. Combining abilities for maize stem antibiosis, yield loss, and yield under infestation and no infestation with pink stem borer attack. Crop Sci. 39:691–696.[Abstract/Free Full Text]
- Butrón, A., N.W. Widstrom, M.E. Snook, and B.R. Wiseman. 2002. Recurrent selection for corn earworm (Lepidoptera: Noctuidae) resistance in three closely related corn southern synthetics. J. Econ. Entomol. 95:458–462.[Web of Science][Medline]
- Cochran, W.G., and G.M. Cox. 1957. Experimental designs. 2nd ed. John Wiley and Sons, Inc., New York.
- Cordero, A., R.A. Malvar, A. Butrón, P. Velasco, P. Revilla, and A. Ordás. 1998. Population dynamics and life-cycle of corn borers in South Atlantic European coast (Lepidoptera, Noctuidae, Pyralidae). Maydica 43:5–12.[Web of Science]
- Cornelius, P.L., J. Crossa, and M.S. Seyedsader. 1996. Statistical test and estimators of multiplicative models for genotype-by-environment interaction. p. 199–234. In M.S. Kang and H.G. Gauch (ed.) Genotype-by-environment interaction. CRC Press, Boca Raton, FL.
- Crossa, J. 1990. Statistical analyses of multilocation trials. Adv. Agron. 44:55–85.
- Crossa, J., and P.L. Cornelius. 1997. Sites regression and shifted multiplicative model clustering of cultivar trial sites under heterogeneity of error variances. Crop Sci. 37:406–415.[Abstract/Free Full Text]
- Crossa, J., P.L. Cornelius, and W. Yan. 2002. Biplots of linear-bilinear models for studying crossover genotype x environment interaction. Crop Sci. 42:619–633.[Abstract/Free Full Text]
- Decoux, G., and J.B. Denis. 1991. INTERA. Logiciels pour l'interprétation statistique de l'interaction entre deux facteurs. Laboratoire de Biométrie, INRA, Route de Saint-Cyr F78026, Versailles, France.
- Denis, J.B. 1980. Analyse de régression factorielle. Biom. Praxim. 20:1–34.
- Denis, J.B. 1988. Two way analysis using covariates. Statistics 19:123–132.
- Dicke, F.F., and W.D. Guthrie. 1988. The most important corn insects. p. 767–867. In G.F. Sprague and J.W. Dudley (ed.) Corn and corn improvement. Agron. Monogr. 18. ASA, CSSA, and SSSA, Madison, WI.
- Epinat-Le Signor, C., S. Dousse, J. Lorgeou, J.B. Denis, R. Bonhomme, P. Carolo, and A. Charcosset. 2001. Interpretation of genotype x environment interactions for early maize hybrids over 12 years. Crop Sci. 41:663–669.[Abstract/Free Full Text]
- Gabriel, K.R. 1971. The biplot graphic display of matrices with application to principal component analysis. Biometrika 58:453–467.[Abstract/Free Full Text]
- Galichet, P.F. 1982. Hibernation d'une population de Sesamia nonagrioides Lef. (Lép. Noctuidae) en France méridionale. Agronomie 2:561–566.
- García-Martí, F., J. Costa Comelles, and F. Pérez. 1996. Las plagas agrícolas. Phytoma 16.
- Hébert, Y., C. Plomion, and N. Harzic. 1995. Genotype x environment interaction for root traits in maize, as analyzed with factorial regression models. Euphytica 81:85–92.
- Hilal, A. 1981. Etude du développement de Sesamia nonagrioides et établissement de modèles pour la prévision de ses populations dans la nature. Bull. OEPP 11:107–112.
- Klenke, J.R., W.A. Russell, W.D. Guthrie, and O.S. Smith. 1988. Inbreeding depression and gene frequency changes for agronomic traits in corn synthetic selected for resistance to European corn borer. J. Agric. Entomol. 5:225–233.
- Larue, P. 1984. La Sésamie du maïs (Sesamia nonagrioides Lef.) dégâts et actualisation de la lutte. La Défense des Végétaux 227:163–179.
- Lynch, R.E. 1980. European corn borer: Yield loss in relation to hybrid and stage of corn development. J. Econ. Entomol. 73:159–164.
- Moreno-González, J. 1988. Variétés de maïs adaptées au nord-ouest de l'Espagne. Phosphore Agric. 80:55–62.
- Nyhus, K.A., W.A. Russell, and W.D. Guthrie. 1989. Changes in agronomic traits associated with recurrent selection in two corn synthetics. Crop Sci. 29:269–275.[Abstract/Free Full Text]
- Revilla, P., R.A. Malvar, M.E. Cartea, A. Butron, and A. Ordás. 2000. Inheritance of cold tolerance at emergence and during early season growth in maize. Crop Sci. 40:1579–1585.[Abstract/Free Full Text]
- Russell, W.A., G.D. Lawrance, and W.D. Guthrie. 1979. Effects of recurrent selection for European corn borer resistance on other agronomic characters in synthetic cultivars of corn. Maydica 24:33–47.
- SAS. 2000. The SAS System. SAS online doc. HTML format version eight. SAS Institute Inc., Cary, NC.
- Van Eeuwijk, F.A., L.C.P. Keizer, and J.J. Bakker. 1995. Linear and bilinear models for the analysis of multi-environment trials: II. An application to data from the Dutch maize variety trials. Euphytica 84:9–22.[Web of Science]
- Yan, W., L.A. Hunt, Q. Sheng, and Z. Szlavnics. 2000. Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Sci. 40:597–605.[Abstract/Free Full Text]
- Vargas, M., J. Crossa, K.D. Sayre, M.P. Reynolds, M.E. Ramirez, and M. Talbot. 1998. Interpreting genotype x environment interaction in wheat by partial least squares regression. Crop Sci. 38:679–689.[Abstract/Free Full Text]
Related articles in Crop Science:
- THIS ISSUE IN CROP SCIENCE
Crop Science 2004 44: 707-710.
[Full Text]
This article has been cited by other articles:
|
|
|
|
 
R. A. Malvar, P. Revilla, A. Butron, B. Gouesnard, A. Boyat, P. Soengas, A. Alvarez, and A. Ordas
Performance of Crosses among French and Spanish Maize Populations across Environments
Crop Sci.,
May 6, 2005;
45(3):
1052 - 1057.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
