Combining Ability for Yield of Single-Cross Hybrids Derived from Maize Composites ( Zea mays L . )

Development of high-yielding hybrids of maize depends on good understanding of combining ability and inheritance of yield trait. To achieve this goal, synthetic populations of lines are developed and improved upon by recurrent selection to be utilized as initial material for the creation of inbred lines. Therefore, the aims of the present paper were: to determine the combining ability among some inbred lines developed from composites and commercial hybrids by recurrent selection method and to choose the most promising inbreds for maize yield improvement and the most suitable hybrid combinations. The highest value of the general combining ability (GCA) was obtain from inbred line TA 447 (1,655 kg ha) followed by tester TC 399 (233 kg ha). For the specific combining ability (SCA), TA 447 × TC 385 A, TA 427 × TC 399, TA 428 × TC 399 and CO 305 × TD 268 had good compatibility. The research concluded that GCA was more important that SCA, indicating that the additive genetic effects are mainly involved in the heredity of yield potential of maize. For the non-additive effects to increase, the next selection cycles must focus on plants’ pairs with more pronounced heterosis for the yield trait.


Introduction
Potential for high maize yield, though it is most complex feature (Sarca, 2004), is the main aim of breeding programs because it makes the greatest contribution to the economic efficiency of a culture.To achieve this, the entire genetic system of the plant acts through its mechanisms on a large number of physically, chemically, biologically, physiologically and biochemically interrelated processes (Sarca, 2004).Increasing yield potential is a combination of genetics and management practices (Hmielowski, 2018).Development of high-yielding hybrids depends on good understanding of combining ability and inheritance of important quantitative traits such as grain yield (Owusu et al., 2017), as well as good choice of the proper inbred lines.According to Hallauer (1990) production potential can only be identified by testing the combining capacity of inbred lines."The objective of the maize breeder should thus not be to find the best pure line, but to find and maintain the best hybrid combination" (Hallauer, 1988).Quantitative genetic traits, such as yield, are characterized by cumulative actions of many factors which include gene effects and effects due to the interaction of genotype and environment (Djemel et al., 2012).Gene effects represent a big interest for plant breeders in order to formulate the most advantageous breeding procedures for improvement of the yield (Gamble, 1961).According to Bauman (1981) a way to improve yield potential is to develop synthetic populations of lines and to improve them by recurrent selection for using them as the initial material for the creation of inbred lines.Based on the information presented above, the aims of the present paper are (i) to estimate the combining ability of some inbred lines developed by recurrent selection method, (ii) to single-out the most promising inbreds that can be used for improving maize yield, and (iii) to determine the most suitable hybrid combinations.

Experimental environment
The experiments were conducted at Agricultural Research and Development Station Turda, Romania.The Research Station is located on Transylvanian Plateau in Romania, with the following coordinates: 46°35' N latitude, 23°47' E longitude, 345-493 m altitude.Climatic Carolina" model (Căbulea, 2004).Grain yield of every hybrid from experimental system was calculated after the formula: where: μ = average yield of experimental hybrids; ĝm = additive effects due parent m; ĝn = the additive effects due parent n; ŝmxn = non-additive effects due to m × n crossing.Correlations, regressions and determination coefficients in the classical way were computed to connect the additive and non-additive genetic effects with yield potential.
The contribution of testers, tested inbred lines and interactions to the total variance had been presented in Fig. 1.The variance of tested inbred lines had the highest value on total variance of yield and variance, due interaction was the lowest.
Analyzing the variability of testers (Table 4), the TC 399 tester was remarked for the general combining ability.The tester lines still need genes accumulations with additive and dominant effects for the more pronounced GCA expression.Among the inbred lines tested for the general combining ability, highly significant differences (***) presented the line TA 447.
conditions are presented in Table 1.Type of predominant soil was Luvic Chernozems (CH lv) (WRB-SR-19981, World reference base for soil resources, 2014) 1 .Seeds were sown using a density of 60,000 plants ha -1 in a non-irrigated system.

Genetic materials
The present research was carried out over two experimental years (2011 and 2012) using 24 single-cross hybrids (21 experimental hybrids and 3 controls).The hybrids were developed through crosses between inbred lines derived from composites populations ('Tu Comp A' and 'Tu Comp B') developed from BSSS and Lancaster Sure Crop germplasm group and from other commercial hybrids (details in Table 2).The crosses were performed at Agricultural Research and Development Station Turda on 2010.

Data analysis
An important aim in breeding programs is to gain knowledge about the genetic factors controlling maize yield and other important characters (Hallauer et al., 2012), thus we employed Analysis of Variance (ANOVA) which allows simultaneously study of the variability due to every factor involved in the yield potential of maize.The variance of genotypes was splited into influence of testers, inbred lines and interactions, using the model described by Haș et al. (2010).Genetic effects were calculate using 2 nd "North 466  Yield potential of simple hybrids ranged from 7 to 10.31 tons ha -1 .Specific combining ability touched values enclosed by -303 and 334 kg ha -1 .
The regression line, the regression equation (y), the correlation coefficient (r) and the determination coefficient (R 2 ) for the studied crossing system and for the correlations of additive and non-additive genetic effects and maize yield are presented in Figs. 2 and 3.The correlation coefficient (r) indicates a strong relation between yield and the sum of additive effects; r was distinctly significant different from zero (**) and the value of R 2 denote that the sum of additive effects strongly influences yield's gain: for every 100 kg ha -1 growth of additive effects, maize yield grew with 95 kg.
The correlation between non-additive genetic effects and yield was not statistically significant (Fig. 3).The bond between the two variables was lower than the sum of additive genetic effects indicating through determination coefficient less dependence of maize yield on non-additive genetic effects.P 5% = 552 kg ha -1 P 1% = 729 kg ha -1 P 0.1% = 939 kg ha -1 Note: Avg: average yield; ĝm, ĝn: additive genetic effects; ŝm×n: non-additive genetic effects; LSD: least significant differences for the experimental trials; P (5%, 1%, 0.01%): probability level Fig. 1.Contribution of different types of variance on total variance of the yield more important than specific combining ability.Anderson et al. (2012) also suggested that the biggest proportion of genetic variation in corn is due to additive genetic effects.Even for tropical and subtropical maize inbred lines, Erdal et al. (2015) concluded that the additive genetic variance is the most important for grain yield.Regarding the contribution of different types of variance on total variance of the yield in the present research, the low variance due to interaction proofed that in the studied inbred lines the contribution of genes with non-additive effect was relatively low compared to the role of additive genetic effects; this situation suggest that for the non-additive effects to increase, the next selection cycles must focus on pairs of plants with more pronounced heterosis for the yield trait.We further concluded from analysis of additive and non-additive genetic effects for the noted hybrids (Table 4), we can conclude that additive gene effects were predominant in producing productive hybrids, both in the case of inbred lines derived from composites and in the case of inbred lines derived from commercial hybrids.Although the association Fig. 2. Regression line and equation for grains' yield according to general combining ability effects (∑ ĝ m +ĝ n ) for the studied experimental cross system Fig. 3. Regression line and equation for grains' yield according to specific combining ability effects (ŝ m×n ) for the studied experimental cross system between additive genetic effects and maize yield or between non-additive effects and the same trait are weak and less compared to the association between additive effects and production capacity, yet these non-additive genetic effects can differentiate between two performing hybrids.For instance, Souza et al. (2009) concluded that for low and high stress environments, the non-additive genetic effects were the most important for developing a promising hybrid.The same results were also confirmed by Murtadha et al. (2018) who asserted that the hybrids with superior specific combining ability indicate that dominance effects are more effective than additive genetic effects in heredity of some yield elements under low water condition.Finally, we can conclude that if two performing inbred lines with high combining ability for yield are crossed and the non-additive genetic effects between them are high and positive, a new high-yielding hybrid has been found.

Conclusions
Our research concluded that: (i) the general combining ability was more important than specific combining ability, (ii) the inbred line TA 447 and the tester TC 399 had the highest general combining ability; (iii) for the specific combining ability several hybrid combinations were good combiners: TA 447 × TC 385 A, TA 427 × TC 399, TA 428 × TC 399 and CO 305 × TD 268; (iv) to increase the role of the non-additive effects, the next selection cycles should focus on the pairs of plants with a more pronounced heterosis for the yield potential.

Table 1 .
Rainfalls and average air temperature during the 2011 and 2012 growing seasons

Table 2 .
Testers, tested inbred lines and control hybrids used in the experimental system

Table 3 .
ANOVA for yield potential on the experimental cross system DF: degree of freedom; s 2 : variance; **significant at the level of probability p = 0.01; *significant at the level probability p = 0.05

Table 4 .
Grains yield, additive and non-additive effects for the maize inbred lines from the experimental cross system