Yield behavior and its components

In Table 3 the H3 and H5 females presented an average grain yield with 8.13 t ha^-1; however, H6, H7, H2 and H1 were in the second statistical group with averages of 8.01, 7.88, 7.84 and 7.54 t ha^-1 respectively; The outstanding yield in females could be due to their intermediate cycle and drought tolerance, in males M2 (8.12 t ha^-1) and M4 (7.9 t ha^-1) stood out. It is probable that the combination between males and females influenced the behavior of the crosses, since the similarity in the yield obtained from the male lines is due to the fact that they are of intermediate cycle, which is reflected in their superior yields to the rest of the males. ^{Wong et al. (2007}) also related this component which allows a high expression with a higher yield. Therefore, a good yield would be expected when combining these parents in the formation of hybrids (^{Soengas et al., 2003}).

For plant height (AP) the females with the highest bearing were H6 (2 m) and H2 (1.99 m), this type of height is ideal for mechanical harvests because they can resist lodging, the females presented on average a height of ear (AM) favorable for mechanical harvests were H3 (1.44 m) and H4 (1.43 m), in female flowering the averages in the seven females fluctuated between 74 to 76 days (d), and they are considered as intermediate cycle , which is an advantage since farmers prefer to sow materials from this cycle, in ear length (ML) the H7 had the largest size with 14.88 cm, while the H5 was in the second statistical group with 14.31 cm, the rest of the females were concentrated in different groups, in ear diameter (DM) females H1, H6 and H3 presented averages of 4.66, 4.31 and 4.23 cm respectively; in rows per ear (HM) and grains per row (GH), there were no statistical differences.

For males there was no statistical significance in (RG), (AP), (DM) and (HM), however, for ear height (AM), M4 presented an average of 1.53 m and M1 obtained the lowest value with 1.29 m, for female flowering, M4 was earlier with 60 d, while the rest of the males obtained averages of 80 d, in ear length and cob diameter there were no marked differences, only 1.2 and 0.23 cm was the range of difference In both variables, respectively, in grains per row, the lowest average was presented in M2 with 30.38 and the highest average was obtained by M3 with 34.24.

By combining, by recombination and selection some contrasting lines, it could result in defined heterotic patterns because, according to ^{Dzib et al. (2011}), the highly yielding hybrids were due to heterosis and also due to hereditary factors, such as the multiple interaction of additive type genes. This shows the convenience of knowing the relative importance of each genetic component in the production of hybrids. Lines H3 and H5 were the ones that intervened in the most outstanding crosses for their yield, followed by H6 and H7. M2 was the one that appeared most frequently in superior crosses, so lines H6 and M2 would be indicated to separate the rest of the lines into two heterotic groups (^{Malacarne and San Vicente, 2003}).

General combinatorial ability (GCA)

The highest general combining ability effects for RG and DM were found in H4 and H2 females (Table 4); however, H5 and H1 obtained lower and negative values, which suggests that H4 and H2 contain genes with additive effects that are favorably expressed in grain yield (Ávila et al., 2009). For AP, H2 and H3 showed higher effects on FF, H6 and H4 obtained higher values, for LM females H1 and H3 indicated that they had higher expression with positive effects, however, for DO, HM and GH, H3 and H5 obtained on average the maximum values, all with positive effects and also with the maximum values per se recorded; that is, on average these populations generated the best combinations.

These results indicate that these populations have a high contribution in the expression of the characteristics evaluated in their progeny, and that the additive effects are the most important; therefore, they can be included in a maize genetic improvement program, to contribute with superior alleles (^{De la Cruz et al., 2005}; ^{Preciado et al., 2005}; ^{Guillen et al., 2009}). It is observed that according to the effects of GCA registered there are other populations with diverse origins that can contribute to the genetic improvement of maize and that until now little attention has been paid to their potential as sources to increase the genetic base of commercial hybrids (Guillen et al., 2009).

The males that expressed the highest combining ability for RG and LM were M2 and M1, in AP, AM, FF, LM, DM, DO, HM and GH, M1, M3 and M4 obtained higher and positive values associated between these variables of GCA, which indicates that not all components of grain yield have similar behavior, but depend on the genetic makeup of each line; in addition, M4 showed higher values in six (60%) of the characteristics, which can be advantageous in improving yield, as mentioned by ^{Wong et al. (2007}).

In all the characters mentioned above, the effects of GCA were greater than those corresponding to SCA, although it is important to note the presence of dominant genes, but the effects of additive genes contribute in a greater proportion than the non-additive effects in the expression of the phenotype, which coincides with ^{Badawy (2013}) for the grain yield variable. However, these results differ from those published by ^{Kanagarasu et al. (2010}); ^{Borghi et al. (2012}) who found that the non-additive effects were higher than the additives for grain yield, ear diameter, ear length and plant height.

Specific combinatorial ability (SCA)

In Table 5, the average value and the effects of SCA of the 28 hybrids are shown, observing that the five best combinations for RG (Table 6) were presented in (H4*M2), (H6*M3), (H2*M3), (H5*M4) and (H6*M4); however, the hybrid (H1*M2) turned out to be the one that expressed the lowest aptitude with a negative value, where the maximum value was 1.17 t ha^-1, which corresponded to the first listed cross, which was expected since the parent lines of H4xM2 expressed the greatest effects of GCA (Table 5), the above, may be due to the sum of the additive effects of the genes of the parents or a high positive effect of SCA, this result suggests that the manifestation of high yield in the crosses is due to the gene action of additive effects of the lines, as asserted by ^{Springer and Stupar (2007}). It was detected that in the crosses with higher magnitude SCA, they combined populations of contrasted geographical origin in the distribution area of the state of Nuevo Leon, and that the dimension of the effects recorded was due to the distance that exists between both populations; H4 belongs to the municipality of Pesqueria and M2 to Los Ramones.

In AP, the best five hybrids that expressed greater SCA effects were: H2xM4, H3xM1, H4*M2, H6*M1 and H5*M2. The populations belonging to the variants of the hybrids had an important influence on the expression of the outstanding crosses for this variable, where said crosses mostly presented both average values and high SCA effects, in AM the best crosses were (H4*M4), (H4*M2), (H5*M2), (H2*M1) and (H7*M4). For FF and LM, the simple crosses H2*M3 and H3*M3, expressed greater effects of SCA; however, in DM, DO, HM and GH, they were H1*M3, H1*M1 and H5*M1.

On the other hand, of the eight parents involved in the crosses with the highest SCA effects, H1 and M1 showed high GCA effects (Table 5), which indicates that their parents are suitable to form intervarietal hybrids with high yield potential, or to derive lines that when crossed have a good combination between them. In this regard, it is expected that crosses with higher SCA result from crossing at least one population with high GCA (^{Legesse et al., 2009}; ^{Escorcia et al., 2010}).

The parents H4, H2, M1 and M2 were the ones that yielded the most (from 7.3 to 8.1 t ha^-1), presented a plant size (from 1.8 to 1.9 m) and ear (from 1.2 to 1.4 m), of a cycle intermediate (from 76 to 81 days at FF), but with greater ear length (13 to 14.3 cm), greater number of rows per ear (12.6 to 13) and greater number of grains per row (30.3 to 33) with respect to the rest of the lines.

This indicates that non-additive genes participated in the inheritance of the crosses. The non-additive action of the genes was reduced between the crosses of the parents, the participation of non-additive genes in the best crosses may be due to the fact that the parents have wide genetic variation and greater additive effects, as corroborated by the high GCA values found. in the variables RG, AP, AM, LM, DM, DO, HM and GH.

These results coincide with that reported by (^{Dzib et al., 2011}), who reported that non-additive genes contributed more to the expression of ear length and diameter, number of rows to obtain the highest grain yield. Crosses with higher SCA values can be considered as prospects in genetic improvement programs to form hybrids and to introduce genetic variation in recurrent reciprocal selection programs. The values were small compared to those obtained by ^{Reyes et al. (2004}); ^{Escorcia et al. (2010}), but similar to those of Legeese et al. (2009).

Genetic components

When estimating the genetic components (Table 6), in grain yield and ear length, the variances of hybrids, males, females and interaction between males by females, high values ​​were observed; however, in the environmental variance the variables such as RG, AP, LM, DO and GH were higher with respect to the additive and dominance variance, so that the environmental variance influenced in a greater proportion the expression of the phenotypic variance with respect to these characters. In the case of heritability in the broad sense, the RG, AM and HM stood out, with the exception of the LM, AM and FF for heritability in the narrow sense.

These results are attributed to the diversity of the genetic makeup of the progenitor lines, which when mating form individuals of a heterozygous type, which reduce the additive effects. For this reason, it is to be assumed that all the variations are estimated based on the genetic behavior determined as dominance variance, which is to be expected since there is a greater heterotic expression in the hybrids due to the effect of the SCA of most of the lines in RG, which results in a higher heritability value in the broad sense due to the effect of the dominance variance (^{Badawy, 2013}). ^{Hallauer et al. (2010}) pointed out that another of the possible reasons for a lower additive variance is the origin of the parents that form the hybrids, since when they come from a single population or related populations, there is a high probability that this type of gene action will stand out, as is the case of kinship between the M2 lines (Population 19 of Liebre, early drought tolerant) and the H4 (Population 11 of Pinto Amarillo, intermediate cycle of wide adaptability) used in this work.

The greatest contribution was found in the environmental variance (σ² F), so it is suggested to form superior hybrids in yield and their components, with the selection of lines that complement each other in their characters, as well as increasing the outstanding lines in GCA and SCA, analyze and select for grain yield, with the intention of releasing new hybrids (^{Preciado et al., 2005}). Therefore, it is suggested to form heterotic synthetic varieties (with 8 to 10 lines), for which lines that complement each other in their characters and that form contrasting populations are selected to recombine and extract new lines to form hybrids superior in yield, in schemes of recurrent reciprocal selection (^{Hallauer et al., 2010}) the heritability values in the strict sense (h²) for DM, DO and GH were 28.14, 29.34 and 29.37%, respectively, coincide with that reported by Molina (1992) who he identified them as quantitative traits of intermediate heritability. The h² of the RG also coincided with the classification of the cited author (low heritability character).

Heritability refers to the ability of characters to be transmitted from generation to generation; that is, it can be considered as the degree of similarity between the individuals of one generation and the next, for which it is considered these would be the characters that would be recombined and selected. The components (LM, DO, DM and HM) that most influenced grain yield, presented high h² (Molina, 1992). This indicates that they are closely associated with grain yield, which is considered high since they are quantitative traits, except for negative heritability variables whose value seems logical as they are more than qualitative traits. These effects may also be due to the fact that the same group of genes does not control all the responses of the genetic components, as pointed out by ^{Springer and Stupar (2007}), or hereditary factors such as additivity, dominance and over-dominance, according to ^{Lippman and Zamir (2007}).