Printer Friendly

Biplot analysis of forage combining ability in maize landraces.

ENSILING MAIZE stabilizes forage availability throughout the year, since it permits a supply of good quality forage when seasonal production is low. Although the whole maize plant (ears plus stover) is used for silage, most hybrids utilized for silage were selected for grain production alone. Several authors have emphasized the importance of improving the vegetative component in silage (Argillier et al., 1995; Barriere and Traineau, 1986; Dhillon et al., 1990), and if this is important, two constraints must be considered. First, most temperate breeding programs are based on the use of the Stiff Stalk x non-Stiff Stalk heterotic pattern within the Corn Belt Dent race, which has already undergone several cycles of selection primarily for improved grain yield. Second, breeding efforts devoted to improving maize for forage production were made mainly with early maturing genotypes not suitable for warm temperate or subtropical regions. Because of these issues, we believe that exotic germplasm should be considered in breeding programs for which the objective is selection of genotypes for forage production.

An appropriate starting point in such a program could be the evaluation of accessions in germplasm banks that have undergone little or no selection for grain production. Many authors have emphasized the importance of broadening the genetic variability on which most maize breeding programs are based (Goodman et al., 1988; Wilkes, 1993). A result of this concern is the Latin American Maize Project (LAMP), which has cooperatively evaluated nearly 15 000 Latin American landraces (LAMR 1991; Taba, 1994; Salhuana and Sevilla, 1995). Even where landraces have been used as a source of inbred lines in the past, their exploitation has generally been limited (Smith, 1988).

Little attention has been paid to the introduction of exotic germplasm for the production of forage maize, despite some promising initial results. Thompson (1968) found that a group of exotic and semi-exotic populations yielded on average 28% more digestible dry matter than adapted hybrids, and Stuber (1986) suggested that some semi-exotic materials might be suitable for silage, given their good grain production and great vegetative development. More recently, Bertoia (2001) noted that landraces with no history of breeding for grain production generated crosses with good forage potential, whereas inbred lines from the North American Corn Belt did not demonstrate potential for enhanced stover yield and quality when compared with inbred lines from Argentine germplasm (Bertoia et al., 2002).

The races from which most current commercial maize lines were developed represent a very restricted sample of the genetic variability available within the species (Hallauer, 1990), and landraces from tropical and subtropical areas may harbor new favorable alleles that are lacking in more elite temperate germplasm. For hybrid breeding programs, evaluation requires determination of both combining ability and per se performance. Abel and Pollak (1991) suggested that the effective value of a germplasm depends on its heterotic response with other genotypes. Thoughtful utilization of local germplasm accessions and the search for alternative heterotic patterns are therefore important research objectives. Exploitation of germplasm from non-Corn Belt Dent racial backgrounds in hybrid breeding programs will be facilitated by the identification of heterotic patterns between germplasm sources.

Diallel crosses have been widely used to determine heterotic responses and heterotic patterns in maize populations (Hallauer and Miranda Filho, 1981), with conventional diallel analysis limited to partitioning total variation into GCA and SCA. Recently, Yan and Hunt (2002) suggested the application of principal component biplot techniques to diallels. In the original application, the first two principal components of the data matrix obtained from a multienvironment trial were used to display genotype effects and genotype x environment interactions in a single two-dimensional figure. By analogy to multi-environment trials, the first two principal components of a diallel data matrix can be extracted and used to display GCA and SCA effects.

Based on results of previous evaluations of germplasm accessions in Argentina, and in some instances in cooperation with LAMP (Ferrer and Solari, personal communication, 1997), eight landraces from Argentina were selected and crossed following a diallel mating scheme. The objectives of this study were: (i) to evaluate the performance of the eight selected landraces, (ii) to determine their genetic potential as sources of germplasm for a forage maize-breeding program, and (iii) to identify heterotic groups among populations with different geographic origins by principal component bi-plot analysis.


Eight maize landraces were selected for their agronomic response, differences in geographic origin, maturity, and grain type (Table 1). Seeds were supplied by the Maize Germplasm Bank at INTA Pergamino, Argentina. Landraces were crossed following a diallel mating design without reciprocals. Crosses were performed in eight isolation blocks. In each isolation block, one population was used as the male and the other seven populations were detasseled and used as females. At least 150 ears per cross were obtained. Landraces per se, the 28 F1 crosses, and four commercial check hybrids (Cargill Semiden 5, Dekalb 4F37, Morgan 369, and Syngenta Pucara, selected for grain production but widely used for forage production in Argentina) were evaluated during two growing seasons (1997-1998 and 1998-1999) at Esteban Echeverria (34[degrees]38' S, 58[degrees]48' W) and Vicente Casares (35[degrees]18' S, 58056' W) in the Buenos Aires Province dairy region. Soils are typical Argiudoll (Vicente Casares) and Aquic Argiudoll with silty clay loam and B2t horizon (Esteban Echeverria), respectively.

The experimental design was a randomized complete block with three replications within each environment. Experimental units consisted of two 5.20-m rows, spaced 0.70 m apart. Plots were over-planted at 52 seeds per row, then thinned to a density equivalent to 71500 plants [ha.sup.-1] at the three-leaf stage. Each experimental unit was harvested by hand when the kernel milk line in approximately 50% of the plants reached two-thirds of the way down the kernels at the central part of the ear (Hunt et al., 1989). Ear and stover were separated and weighed fresh. A representative sample of each plant component was taken, weighed fresh, and dried with dry forced air, then weighed dry to permit estimation of dry matter percentage. Dried samples were milled to 1-mm particle size and analyzed with near-infrared reflectance spectroscopy. Near infrared spectra between 1100 and 2500 nm at every 2 nm were collected on all milled samples using an NIRS 6500 spectrophotometer (NIRSystem Inc., Silver Spring, MD). In vitro dry matter digestibility of ear (ED) and stover (SD) were predicted by NIRS equations, which were calibrated by the enzymatic method (Gabrielsen, 1986). Whole-plant dry matter digestibility (WD) was computed as the sum of ED and SD.

Statistical Analyses and Mathematical Model for GGE Biplot

Analyses of variance were performed for each variable, using a mixed model (McIntosh, 1983), where replications, environments, and genotype x environment interactions were considered random effects.

Applying GGE biplot methods to diallel data, the terms "average yield" and "stability" of the genotypes correspond to GCA and SCA, respectively. The mean values for hybrids and parental populations across environments are used to form a symmetrical diallel data matrix from which the first two principal components are extracted. In this matrix, each population corresponds to one row and one column of data, where the row is considered an "entry" and the column a "tester" (Yan and Hunt, 2002). Thus, each population can be considered both an entry and a tester. Means of each column are calculated and a new, adjusted (nonsymmetrical) data matrix is obtained by subtracting the column (tester) mean from each cell. After obtaining the first two principal components of the adjusted data matrix, the biplot model can be written as:

[[gamma].sub.ij] - [[beta].sub.j] = [[lambda].sub.1][[xi].sub.i2][[eta].sub.j2] + [[epsilon].sub.ij]

where [[gamma].sub.ij] is the genotypic value of the cross between entry i and tester j for the trait of the interest; 131. is the mean of all crosses involving tester j; [[lambda].sub.1] and [[lambda].sub.2] are the singular values for the first and second principal components (PC1 and PC2 respectively); [[xi].sub.i1] and [[xi].sub.i2] are the PC1 and PC2 eigenvectors, respectively, for entry i; [[eta].sub.j1] and [[eta].sub.2] are the PC1 and PC2 eigenvectors, respectively, for tester j; and [[epsilon].sub.ij] is the residual of the model associated with the combination of entry i and tester j. When i [not equal to] j, the genotype is a population hybrid. When i = j, the genotype is a landrace.

For each trait, we obtained the eigenvectors of the first two PCs for each genotype, and the eigenvalues for PC1 and PC2. The singular value for a PC was obtained as the square root of the sum of squares explained by the PC, which is the product of the eigenvalue multiplied by the number of landraces. Principal components scores for entries and testers were scaled symmetrically.

Following Yan and Hunt (2002), we displayed the results of the principal components analysis in two ways: as an average tester coordinate view and as a polygon view. The average tester coordinate (ATC) view is displayed by defining the average tester position in the biplot as that position with average PC1 and PC2 scores of all testers. The ATC is established with its abscissa passing through the origin and the average population, and its ordinate passing through the origin and perpendicular to the abscissa. The GCA effects of the populations are then approximated by their projections as entries onto the ATC abscissa. Grid lines perpendicular to the ATC abscissa are displayed to help to rank the populations in terms of GCA. Projections of the populations as testers onto the ATC ordinate approximate their SCA effects, which represent the trend of the populations to produce superior hybrids with specific genotypes.

The polygon view of a biplot provides an alternative graphical presentation that permits identification of interactions between populations. Polygon view biplots were drawn, resulting in the partitioning of the biplot into sectors, with entries farthest from the center of the biplot representing the vertices of the polygon. Testers within a sector form the best hybrids with the entry at the vertex of the sector. Entries located near the biplot origin are less responsive to the change of testers.

All statistical analyses were performed with the Proc Princomp and Proc Mixed procedures of the SAS software package (SAS Institute, 1999).


Genotypes varied significantly (P < 0.01) for all yield traits (SY, EY, and WY), but not for digestibility traits (SD, ED, and WD) (Table 2). Significant variation among parents (P < 0.01) and among F1s (P < 0.01) was also observed for all yield traits. The difference between check hybrids and experimental genotypes was significant for SY (P < 0.01) and EY (P < 0.01). Stover dry matter yield of some landraces and crosses was significantly greater than the best checks, but no landraces or crosses had EY as high as the two best checks (Tables 3 and 4). On average, checks had greater EY but lower SY than the unimproved genotypes (Tables 3, 4, and 5). As WY is the sum of SY and EY, the differences between these two traits counteracted each other when WY is considered, resulting in no significant difference between checks and unimproved genotypes (Tables 2 and 5). Variation among checks was observed for SY and EY (P < 0.01), but not for WY. Crosses had greater values than parental landraces (P < 0.01) for all yield traits, indicating that heterosis was significant. The GCA effects were significant (P < 0.01) for SY, EY, and WY, but SCA was not significant for any trait.

Biplot Analysis

Ear Yield

The first two principal components explained 67% (42.1 and 24.9% by PC1 and PC2, respectively) of the variation for EY. The average tester coordinate biplot indicates that landraces B, G, D, and F had positive GCA effects (order also indicates ranking order), whereas landraces E, A, H, and C had negative GCA effects (Fig. 1). Although landraces E and H combined well with landraces A and C, their SCA effects were not significant (Table 2).


The biplot in Fig. 1 is divided into five sectors with landraces A, E, H, D, G, and B as the vertex. Two well differentiated and opposite groups can be observed, E to H and A to C. Landraces A, E, H, and C produced the worst combinations with themselves, since testers e, h, a, and c fell into opposite sectors. Testers d, e, f, g, and h fell into sector B to G, indicating that their crosses with B, G, and F generated good hybrid combinations. Sector D contained three adequate hybrid combinations: b x D, a x D, and c x D. High values of MPH for these crosses were observed (Table 3). Both b x G and d x G represent crosses with interesting potential as breeding materials, since they did not have significant differences with the second highest yielding genotype (Cargill Semiden 5), and showed very high MPH (29 and 24%, respectively). Another promising cross is a x D with an EY similar to Syngenta Pucar4 and Morgan 369 and a MPH of 26%.

Stover Yield

Principal components 1 and 2 together explained 87.5% of the observed variation for SY (76.2% and 11.3%, respectively, for PC1 and PC2). The corresponding biplot (Fig. 2) shows that landraces A, B, C, and D had positive GCA effects for SY. We suggest that maturity has a strong influence on SY, since only later flowering landraces ([summation]T10[degrees]C > 800[degrees]C) showed positive GCA effects. Consequently, it is possible to observe a clear separation between landraces with positive and negative GCA effects for SY. Landraces A, B, C, and D exhibited good combining ability with all testers including themselves. The polygon view biplot demonstrates that landraces A and B produced good hybrid combinations with c, d, e, g, and h for SY (sector A-B, Fig. 2). Similarly, it can be observed in sector D that landraces C and D possess good combining ability with testers a and f. The distance between the x axis and a genotype in the biplot is an estimation of its SCA effect (analogous to stability in GGE biplots). All landraces clustered near the x axis, indicating that SCA effects were not important for SY. The entries most distant from the x axis were populations C and D, but their distances from the x axis were nonsignificant. This observation agrees with the ANOVA, where variation due to SCA effects was not significant (Table 2). According to the biplot view, as a is in sector C-D, and c and d are in sector A-B, crosses a x C and a x D are predicted to be the best combinations.


Crosses a x (C, E, G), b x (C, D, E, G), c x E, and d x (E, F, G) did not show significant differences with the best commercial hybrid (Morgan 369), while a x B and a x D yielded greater than this check (Table 4). These two crosses also showed significant MPH values (10 and 17%, respectively).

Whole-Plant Yield

The first two principal components together explained 81.1% (66.3 and 14.8%, respectively) of the variation for WY. The resulting biplot revealed a response similar to that observed for SY (Fig. 3), where the later flowering populations (A, B, C, and D) had positive GCA effects. Based on the distance between each entry and the ATC abscissa, landrace B had the greatest GCA effect, followed by landraces D, A, and C, respectively. Negative GCA effects (in decreasing order) were observed in landraces H, E, F, and G. The polygon biplot view reveals four well-defined sectors, named A, B, D, and H. SCA values for this trait were small and nonsignificant in the ANOVA (Table 2). Consequently, it was not possible to observe clear heterotic groups. Good hybrid combinations were d x A, b x A, and c x A in sector A; h x B, g x B, and e x B in sector B; and f x D and f x C in sector D. Sector H included landraces E, F, and H, but no testers fell into this sector, indicating that these landraces exhibited poor performance for WY both per se and in hybrid combinations. Landrace G did not fit in a well-defined sector, since it could be assigned to sector A or H. In sector A, d was predicted to be the best mating partner for A, while in sector D, a is the best partner for D. Both A and D were, therefore, identified to be the best partners for each other. This combination significantly out yielded all commercial hybrids (Table 5) and showed an MPH of 20%. Crosses b x G and d x G did not show WY differences with the two best checks, Morgan 369 and Cargill Semiden 5.



No significant breeding effort to improve corn forage yield or quality attributes has been undertaken by corn breeders (Lauer et al., 2001), and silage maize currently lacks both well-defined heterotic groups and an appropriate ideotype that can be used to guide forage maize breeders. Breeding of highly digestible forage maize may depend on the re-evaluation and use of old genetic resources (e.g., landraces) that are not currently used or that were never used in maize breeding (Barrirre et al., 2005). Selection for modern grain type hybrids changed plant design (Evans and Fischer, 1999), with an improvement in whole plant digestibility as a result of the increased contribution of EY to WY in commercial checks. However, such improvement was not observed in this study, since WD of commercial hybrids and landraces were not significantly different (Table 2). According to Roth et al. (1970), Gunn (1975), and Twumasi-Afriyie and Hunter (1982), this result may be explained by the hypothesis that selection not only resulted in an increase in harvest index and stalk lodging resistance in modern grain type hybrids, but also in a decrease in SD. Again, however, this effect was not observed in this study, given that there were no differences between landraces and checks for SD. Argillier et al. (1995) reported similar results. According to Lauer et al. (2001), little change has occurred in the quality of the stover portion of corn forage cultivars available in northern corn belt from 1930 to the present.

Biplot analysis was preferred over conventional diallel analysis because its graphical representation jointly gave information about GCA and SCA effects of the populations, their performance in crosses, as well as grouping patterns of similar genotypes. The heterosis values observed and the performance of crosses compared to the hybrid checks, suggest that some populations have potential as breeding material to select genotypes with improved forage production. More than one breeding strategy could be adopted in breeding for improved corn silage, but in all cases, a recurrent selection scheme is recommended to increase the probability of obtaining competitive inbred lines. Two heterotic patterns can be defined: (i) one white dent x white dent grain type, with landraces B and D as a group and landrace A as the other group; and (ii) one white dent x orange flint with B and D as a group and G as the other. Landraces B and D can be intercrossed to form a composite and reciprocal recurrent selection can be implemented with it and landraces A or G. Although we believe that an interpopulation scheme is more appropriate, intrapopulation recurrent selection also can be adopted in the composite and landraces A and G. Another potential strategy would be to cross the selected landraces with inbred lines from well-known heterotic patterns like Reid x Lancaster, to study their performance in hybrid populations. Thus, a group of outstanding lines can be crossed with composite B x D, and the best combination identified. The selected inbred line can then be used as the tester in a recurrent selection program and/or in the selection of new inbred lines from the composite, taking advantage of (i) the normally observed interracial heterosis, and (ii) the use of a line with good per se yield and stalk quality as the female parent of the potential hybrids. At the present time, landraces B and D are being intercrossed to form composite B-D.


Sincere thanks are extended to Drs. Jim Holland, North Carolina State University, and Peter Graham, University of Minnesota, for valuable critiques and suggestions of previous versions of this paper. We also thank Dr. Marcelo Ferrer, Germplasm Bank at INTA Pergamino, Argentina, for supplying seeds of the evaluated Landraces.


Abel, B.C., and L. Pollak. 1991. Rank comparison of unadapted maize populations by testers and per se evaluation. Crop Sci. 31:650-656.

Argillier, O., Y. Hebert, and Y. Barriere. 1995. Relationships between biomass yield, grain production, lodging suceptibility, and feeding value in silage maize. Maydica 40:125-136.

Barriere, Y., D. Dalber, O. Dolstra, C. Lapierre, M. Motto, A. Ordas, J. Van Waes, L. Vlasminkel, C. Welcker, and J.P. Monod. 2005. Past and prospects of forage maize breeding in Europe. I. The grass cell wall as a basis of genetic variation 2nd future improvements in feeding value. Maydica 50:259-274.

Barriere, Y., and R. Traineau. 1986. Characterisation of silage maize: Patterns of dry matter production, LAI evolution and feeding value in late and early genotypes, p. 131-136. In O. Dolstra and P. Miedema (ed.) Breeding of silage maize. Proc. 13th Congress on the Maize and Sorghum Section of EUCARPIA, Wageningen, The Netherlands. 9-12 Sept. 1985. PUDOC, Wageningen, the Netherlands.

Bertoia, L.M. 2001. Forage yield and quality combining ability of maize composites with different improvement levels. Maydica 46: 87-92.

Bertoia, L.M., R. Burak, and M. Torrecillas. 2002. Identifying inbred lines capable of improving ear and stover yield and quality of superior silage maize hybrids. Crop Sci. 42:365-372.

Dhillon, B.S., EA. Gurrath, E. Zimmer, M. Wermke, W.G. Pollmer, and D. Klein. 1990. Analysis of diallel crosses of maize for variation and covariation in agronomic traits at silage and grain harvest. Maydica 35:297-302.

Evans, L.T., and R.A. Fischer. 1999. Yield potential: Its definition, measurement, and significance. Crop Sci. 39:1544-1551.

Gabrielsen, B.C. 1986. Evaluation of marketed cellulases for activity and capacity to degrade forage. Agron. J. 78:838-842.

Goodman, M.M., EC. Gonzalez, and R.N. Holley. 1988. U.S. maize germplasm: Origins, limitations, and alternatives, p. 130-148. In N. Russell and G.M. Listman (ed.) Recent advances in the conservation and utilization of genetic resources. Proc. of the Global Maize Germplasm Workshop, CIMMYT, Mexico. 6-12 Mar. 1988. CIMMYT, Mexico, D.F.

Gunn, R.E. 1975. Breeding maize for forage production, p. 37-58. In Proc. 8th Congress of the Maize and Sorghum Section of Eucarpia, Paris-Versailles.

Hallauer, H.A. 1990. Methods used in developing maize inbreds. Maydica 35:1-16.

Hallauer, H.A., and J.B. Miranda Filho. 1981. Quantitative genetics in maize breeding. Iowa State Univ. Press, Ames, IA.

Hunt, C.W., W. Kezar, and R. Vinands. 1989. Yield, chemical composition and ruminal fermentability of corn whole-plant, ear and stover as affected by maturity. J. Prod. Agric. 2:357-361.

LAMP. 1991. Catalogo del germoplasma de maiz. Tomo 1 and 2. (In Spanish and English.) AARS Special Publication, Beltsville, MD.

Lauer, J.G., J.G. Coors, and P.J. Flannery. 2001. Forage yield and quality of corn cultivars developed in different eras. Crop Sci. 41: 1449-1455.

McIntosh, M.S. 1983. Analysis of combined experiments. Agron. J. 75:153-155.

Roth, L.S., G.C. Marten, W.A. Compton, and D.D. Stuthman. 1970. Genetic variation of quality traits in maize (Zea mays L.) forage. Crop Sci. 10:365-367.

Salhuana, W., and R. Sevilla (ed.). 1995. Latin American Maize Project (LAMP). Stage 4 results from homologous areas 1 and 5. ARS Special Publication, Beltsville, MD.

SAS Institute. 1999. SAS/STAT. Software version 8. SAS Inst. Inc., Cary, NC.

Smith, J.S.C. 1988. Diversity of United States hybrid maize germplasm: Isozymic and chromatographic evidence. Crop Sci. 28:63-69.

Stuber, C.W. 1986. Use of exotic sources of germplasm for maize improvement, p. 19-31. In O. Dolstra and P. Miedema (ed.) Breeding of silage maize. Proc. Congr. Maize Sorghum Section EUCARPIA, 13th, Wageningen, the Netherlands. 9-12 Sept. 1985. PUDOC, Wageningen, the Netherlands.

Taba, S. 1994. The future needs and activities, p. 52-58. In The CIMMYT Maize Germplasm Bank genetic resources preservation, regeneration, maintenance, and use. Maize Program special report. CIMMYT, Mexico, D.F.

Thompson, D.L. 1968. Silage yield of exotic corn. Agron. J. 60:579-581.

Twumasi-Afriyie, S., and R.B. Hunter. 1982. Evaluation of quantitative methods for determining stalk quality in short-season corn genotypes. Can. J. Plant Sci. 62:55-60.

Wilkes, G. 1993. Germplasm collections: Their use, potential, social responsibility, and genetic vulnerabillity, p. 445-490. In D.R. Buxton et al. (ed.). International Crop Science I. CSSA, Madison, WI.

Yan, W., and L.A. Hunt. 2002. Bipot analysis of diallel data. Crop Sci. 42:21-30.

Abbreviations: ATC, average tester coordinate; ED, in vitro digestibility of ear dry matter; EY, ear dry matter yield; GCA, general combining ability; GGE, genotype main effect plus genotype x environment interaction; INTA, Instituto Nacional de Tecnologia Agropecuaria; LAMR Latin American Maize Project; MPH, midparent heterosis; NIRS, near infrared reflectance spectroscopy; PC, principal component; PC1, first principal component; PC2, second principal component; SCA, specific combining ability; SD, in vitro digestibility of stover dry matter; SY, stover dry matter yield; WD, in vitro digestibility of whole-plant dry matter; WY, whole-plant dry matter yield.

Luis Bertoia, * Cesar Lopez, and Ruggero Burak

L. Bertoia, C. Lopez, and Ruggero Burak, Dep. of Agronomy, Universidad Nacional de Lomas de Zamora, Camino de Cintura km. 2, (1832) Lomas de Zamora, Prov. de Buenos Aires, Argentina. This research was supported by the Dep. of Agronomy, Facultad de Ciencias Agrarias, Universidad Nacional de Lomas de Zamora. Received 27 Sept. 2005. * Corresponding author (
Table 1. Population name, germplasm bank code, geographic origin, race,
grain type, and plant height of maize populations evaluated in diallel
crosses in two environments during 1997-1998 and 1998-1999 growing

                                          Geographic origin

        Argentine germplasm    USDA PI        Argentine
Name         bank code         number          province

A           ARZM 17-034        493020        San Luis
B           ARZM 03-056        491799        Entre Rios
C           ARZM 01-150        491741        Buenos Aires
D           ARZM 03-054        491797        Entre Rios
E           ARZM 16-062        516063        Mendoza
F           ARZM 16-042        516043        Mendoza
G           ARZM 19-006        516078        Neuquen
H           ARZM 01-088        491682        Buenos Aires

        Geographic origin

Name              Lat. and long.                      Race

A       32[degrees]13' S, 65[degrees]53' W    Dentado Blanco
B       29[degrees]21' S, 59[degrees]59' W    Dentado Blanco
C       37[degrees]11' S, 62[degrees]45' W    Dentado Blanco Rugoso
D       29[degrees]24' S, 59[degrees]41' W    Dentado Blanco
E       34[degrees]55' S, 67[degrees]32' W    Dentado Amarillo
F       33[degrees]47' S, 69[degrees]03' W    Cristalino Colorado
G       37[degrees]04' S, 69[degrees]09' W    Cristalino Colorado
H       38[degrees]06' S, 62[degrees]14' W    Cristalino Colorado

Name     Grain type     GDD ([dagger])    height


A        white dent          906           2.64
B        white dent          897           2.59
C        white dent          865           2.60
D        white dent          858           2.39
E       yellow dent          795           2.30
F       orange flint         752           2.26
G       orange flint         683           2.19
H       orange flint         666           2.25

([dagger]) GDD, growing degree days between planting to 50% silking, in
[degrees]C, calculated as [summation][(T[degrees]max + T[degrees]min)
x [2.sup.-1]] - 10[degrees]C] and T[degrees]max [less than or equal to]

Table 2. Analysis of variance for stover (SY), ear (EY), and
whole-plant (WY), and in vitro digestibility of stover (SD), and
whole-plant (WD), of eight maize landraces, 28 single-crosses and four
commercial checks evaluated in two environments during the 1997-1998,
1998-1999 growing seasons.

Source                               df        SY          EY

                                                 Mean squares

Environments                           3    235.2 **    307.0 **
Genotypes                             35     30.9 **      6.4 **
Parents                                7     49.3 **      5.3 **
F1                                    27     25.3 **      3.6 *
GCA                                    7     86.4 **     10.3 **
SCA                                   20      3.9         1.3
Parents vs. F1                         1     31.9 **     90.4 **
Checks                                 3     24.6 **     14.4 **
Checks vs. genotypes                   1    983.3 **     77.7 **
Genotype x environment               105      3.5 **      2.5 **
Parents x environment                 21      4.6 **      3.6 **
F1 x environment                      81      3.1 *       2.1 **
GCA x environment                     21      3.6 **      3.4 **
SCA x environment                     60      2.9 **      1.7 **
(Parents vs. F1) x environment         3      6.2 **      3.8 **
Checks x environment                   9      4.6 **      1.3
Checks vs. genotype x environment      3     11.8 **      1.0
Replications (environment)             8     21.2 **      7.9 **
Pooled error                         312      1.2         0.9

Source                                  WY          SD          ED

                                               Mean squares

Environments                         845 **      521.6 **    1914.4 **
Genotypes                             43 **       12.2         19.7
Parents                               60 **       18.4         32.4
F1                                    30 **       11.0         16.9
GCA                                   99 **       13.7         42.8
SCA                                    6          10.0          7.8
Parents vs. F1                        28 **        1.13         6.4
Checks                                 6          11.4         29.5
Checks vs. genotypes                   3          10.8          9.8
Genotype x environment                 7 **       16.0 *       20.1 *
Parents x environment                 12 **       12.0         19.2
F1 x environment                       6 **       16.5 *       20.5
GCA x environment                      9 **       11.4         27.6
SCA x environment                      5 **       18.3 **      18.2
(Parents vs. F1) x environment        15 **       29.2         13.5
Checks x environment                   8 **        2.8         15.5
Checks vs. genotype x environment     20 **       30.5         18.0
Replications (environment)            48 **       26.2 *      124.6 *
Pooled error                          30          11.6         16.9

Source                                    WD

                                     Mean squares

Environments                           996.1 **
Genotypes                               11.1
Parents                                 20.1
F1                                       8.4
GCA                                     19.5
SCA                                      4.5
Parents vs. F1                          21.4
Checks                                  13.6
Checks vs. genotypes                    92.0
Genotype x environment                  11.6 **
Parents x environment                   10.9 *
F1 x environment                        11.7 **
GCA x environment                       13.8 *
SCA x environment                       11.0 *
(Parents vs. F1) x environment          16.1
Checks x environment                    13.0
Checks vs. genotype x environment       56.3 **
Replications (environment)              33.8 **
Pooled error                             7.3

* Significant at the 0.05 probability level.

** Significant at the 0.01 probability level.

Table 3. Mean ear dry matter yield (EY) across four environments for
landraces per se (diagonal), F1 hybrids among landraces (above
diagonal), and commercial checks. Percentage of midparent heterosis is
shown below the diagonal.

([dagger])       A                B          C                  D

A             4826             6932       5532               7100
B             26.3 **          6148       6679               7188
C              6.4 **          14.0 **    5574               6598
D             26.6 **          14.7 **    10.3 **            6391
E             25.4 **          23.9 **    29.2 **            12.8 **
F             18.5 **          20.6 **    19.4 **            11.8 **
G             23.9 **          29.2 **    18.0 **            23.9 **
H             25.6 **          25.5 **    20.9 **            11.6 **
Checks               Morgan 369                  C. Semiden 5
                        6454                         8449
LSD (0.05)

([dagger])       E                    F          G                H

A             5945                  6490       6798              6503
B             6691                  7400       7940              7142
C             6646                  6982       6917              6531
D             6232                  6994       7767              6487
E             4657                  6297       6919              5564
F             16.8 **               6123       7153              6898
G             28.1 **               16.6 **    6145              6511
H             12.5 **               21.5 **    148 **            5232
Checks               Syngenta Pucara                 Dekalb 4F37
                          7498                            8935
LSD (0.05)                                                      752.1

* Significant at the 0.05 probability level.

** Significant at the 0.01 probability level.

([dagger]) A, ARZM 17-034; B, ARZM 03-056; C, ARZM 01-150; D, ARZM
03-054; E, ARZM 16-062; F, ARZM 16-042; G, ARZM 19-006; H, ARZM 01-088.

Table 4. Mean stover dry matter yield (SY) across four environments of
landraces per se (diagonal), F1 hybrids among landraces (above
diagonal), and commercial checks. Percentage of midparent heterosis is
shown below the diagonal.

([dagger])       A                B          C                  D

A            11398            12883       12634               13229
B             10.3 **         11962       12580               11400
C             15.7 **          12.3 **    10445               10658
D             17.4 **          -1.3 ns     -1.3 ns            11142
E             10.5 **           6.2 **     14.0 **              8.4 **
F              3.4 *           -8.2 **      8.0 **             12.3 **
G             14.2 **           7.6 **     10.0 **             16.3 **
H             11.3 **           8.3 **     19.4 **              6.1 **
Checks               Morgan 369                   C. Semiden 5
                       11894                         9643
LSD (0.05)

([dagger])       E                    F          G                  H

A            11624                 10453       11461               9493
B            11471                  9534       11101               9540
C            11454                 10401       10513               9617
D            11268                 11208       11522               8916
E             9644                  9256        9576               8723
F              0.3 ns               8820        9353               7966
G              4.6 **                6.9 **     8673               7678
H             14.0 **               10.0 **      7.1 **            5660
Checks               Syngenta Pucara                   Dekalb 4F37
                           9297                            8578
LSD (0.05)                                                        893.6

* Significant at the 0.05 probability level.

** Significant at the 0.01 probability level.

([dagger]) A, ARZM 17-034; B, ARZM 03-056; C, ARZM 01-150; D, ARZM
03-054; E, ARZM 16-062; F, ARZM 16-042; G, ARZM 19-006; H, ARZM 01-088.

Table 5. Mean whole plant dry matter yield (WY) across four
environments of landraces per se (diagonal), F1 hybrids among landraces
(above diagonal), and commercial checks. Percentage of midparent
heterosis is shown below the diagonal.


Landraces       A                 B         C                    D
A            16224             19814       18166               20329
B             15.4 **          18110       19259               18588
C             12.7 **           12.9 **   161119               17256
D             20.4 **            4.3 **      2.9ns             17533
E             15.1 **           12.1 **     18.1 **              9.9 **
F              8.7 **            2.5ns      12.3 **             12.1 **
G             17.6 **           15.6 **     13.0 **             19.2 **
H             18.0 **           14.4 **     20.0 **              8.4 **
Checks               Morgan 369                    C. Semiden 5
                       18348                          18092
LSD (0.05)


Landraces       E                     F         G                   H
A            17569                  16943      18259              15997
B            18163                  16935      19040              16592
C            17901                  17384      17430              16148
D            17500                  18202      19289              15403
E            14301                  15553      16495              14288
F              6.4 *                14943      16506              14865
G             13.3 **                10.9 **   14818              14189
H             13.4 **                15.1 **    10.4 **           10893
Checks               Syngenta Pilcara                  Dekalb 4F37
                         16794                             17513
LSD (0.05)                                                       1386.5

* Significant at the 0.05 probability level.

** Significant at the 0.01 probability level.

([dagger]) A, ARZM 17-034: B, ARZM 0,;-056: C, ARZM 01-150: D, ARZM
03-054: E, ARZM 16--062: E ARZM 16-{142: G, ARZM 19-006: H, ARZM
COPYRIGHT 2006 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bertoia, Luis; Lopez, Cesar; Burak, Ruggero
Publication:Crop Science
Geographic Code:1USA
Date:May 1, 2006
Previous Article:Planting density influences disease incidence and severity of Sclerotinia blight in peanut.
Next Article:Elevated atmospheric carbon dioxide and weed populations in glyphosate treated soybean.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters