Introduction

In 2012, 7.3 million ha of corn were sown and production of 22.07 million (3.19 t ha^-1) was obtained, with 7.5 and 2.2 t ha^-1, irrigated and temporary, respectively (^{SIAP, 2014}). In the High Valleys of central Mexico (2 101 to 2 800 meters above sea level), 1 108 267 ha are sown; 134 082 ha in irrigation, 306 828 ha in favorable weather and 667 357 ha in limiting weather (^{Turrent, 1994}).

Since 1950, INIFAP has released more than 40 hybrids and improved varieties (^{Gámez et al., 1996}) formed with S₁-S₄ lines derived from creoles and some backcrosses, but with more inbreeding depression caused by the high frequency of deleterious genes (^{Márquez, 1988}). Recently, S₇-S₈ lines have been derived from simple crosses formed with lines from High Valleys and donors from other regions (^{Velázquez et al., 2013}), these have been selected per se but their combinatorial aptitude has not been evaluated.

The formation and evaluation of mestizos is important to select the best using suitable testers and based on their combinatorial aptitude (^{Welcker et al., 2005}; ^{Lorenz et al., 2009}) and this has been the main method (^{Bernardo, 2001}) to select lines that convey desirable characteristics. Varieties, recycled lines, mixtures of varieties or hybrids have been used (^{Pfarr and Lamkey, 1992}). A good tester should allow to classify the merit of each line and maximize the genetic gain (^{Russell et al., 1992}; ^{Menz et al., 1999}). In the previous context, the objective of this work was to analyze mestizos formed with two simple crosses of CIMMYT to identify the most outstanding.

Materials and methods

Results and discussion

In this study, significant differences were detected (p= 0.05 or 0.01) among localities in the 14 variables. This fact suggests that the environmental heterogeneity that exists in central Mexico forces the rescobcher to establish trials in contrasting sites to identify the most favorable ones (Table 2) ^{Reynoso et al. (2014)} and ^{Torres et al. (2011}, ²⁰¹⁷) commented on differences in altitude and types of climate and soil are the most important (Table 1).

The significant effects that were observed between treatments (p= 0.01) are explained by the statistical differences between mestizos and commercial hybrids (Table 2). ^{Obaidi et al. (1998)}; ^{Castañon et al. (1998)}; ^{Mihaljevic et al. (2005)} observed similar results. This fact evidences an outstanding fraction of the germplasm that exists in the central region of Mexico of the Conic and Chalqueño breeds and of other CIMMYT breeds (^{González et al., 2008}; ^{Reynoso et al., 2014}; Torres et al., 2001, ²⁰¹⁷).

The significant effects that were detected in all the variables between females or between males explain the phenotypic variability that was registered among mestizos (Table 2), the lines evaluated had a different behavior in their respective mestizos and there is genetic diversity (^{Mosa, 2010)}. Both females belong to CIMMYT and the 43 lines are from INIFAP, derived from Michoacan 21, Cuatero of the Virgen and Tlaxcala 151. ^{Castellanos et al. (1998)} concluded that simple crosses were the best alternative as testers to generate superior trilincob hybrids. In other studies, it was concluded that the behavior of the males was different, depending on the female used (^{Mosa et al., 2008}; ^{Mosa, 2010}; ^{Habliza and Khalifa, 2015}).

The 10 commercial hybrids were statistically different in 10 of the 14 variables. The origin of three of them is unknown, but it is inferred that they could have CIMMYT germplasm, such as H-40 and ICAMEX 2010; the female of the first has the same lines of the tester 1 and the female of the second has the same progenitors of the tester 2. H-58E, H-70, H-66, H-76E and H-77E are formed with germplasm of CIMMYT and with lines of Michoacán 21 and Tlaxcala 151, conic race. In other studies, similar inferences have been made (^{González et al., 2008}; ^{Quiroz et al., 2017}). This last situation could explain the cancellation of effects that occurred in crosses vs hybrids; the mean of each group and its average could be considered as an estimator of the mean of the 96 treatments.

The interaction treatments x localities significant in 11 variables (p= 0.05 or 0.01) suggests that crosses and hybrids are unstable. In the rest of the interactions presented in Table 2, a similar trend was observed. This condition will make it difficult to identify high-performance genetic material with adaptability to the environmental conditions of the study area. Phenotypic instability forces the plant breeder to choose genotypes with specific adaptation; the generation, validation, application and/or transfer of technology will also be conditioned and there will be side effects in the seed increase and production programs. Mexico could be self-sufficient in corn production by identifying a higher fraction of the germplasm available in different rescobch and teaching institutions (^{González et al., 2010}; ^{Franco et al., 2016}; ^{Torres et al., 2011}, ²⁰¹⁷).

The highest grain yields were observed in Valley Toluca-Atlacomulco (from 6 635 to 8 700 kg ha^-1). The greatest biological cycle was also registered, the biggest and the best aspects of plant and cob heights, less acame stem, root and both, and more prolific. Cob rot was less than 6% and there was more tillering (Table 3). Other studies have highlighted the high potential of this region of Mexico, where there are good climate and soil conditions that favor the growth and development of the maize, in the absence of late or cobly frosts and low rainfall; between 4.0 and 10 t ha^-1 of grain have been recorded (^{González et al., 2008}; ^{González et al., 2010}; ^{Reynoso et al., 2014}; ^{Quiroz et al., 2017}; ^{Torres et al., 2011}, ²⁰¹⁷).

In relation to the 96 treatments, the grain yield varied from 3 176 to 7 312 kg ha^-1; 72 mestizos (83%) were statistically equal to H-40, of these, 33 (38%) exceeded it from 0.5 to 15.8% and 23 or 20 were statistically more precocious in DFM or in DFF and had similar means in ALP, ALM, ASM, PMC, PHI and PMP. Only 41 and 84 presented worse ASP, attributable to their higher PAR, PAT and PACA. Crossing 70 was for PAT and PACA. Crosses 38 and 81 presented higher PPC. In this region of Mexico, hybrids and varieties of high yield and stability, cobly, intermediate plant and cob heights, resistant to lodging are desirable (^{González et al., 2008}; ^{González et al., 2010}; ^{Torres et al., 2011}, ²⁰¹⁷), such as those identified as 81, 34, 41, 10, 33 and 59 (Table 4).

The female one (CML246 x CML242) excelled in grain yield, plant and cob heights, tillering and plant and cob aspects and had lower averages in male and female blooms, root, stem and total lodging, cob cover, cob rot and prolificacy (Table 5). Its superiority has been highlighted in other hybrids of High Valley in Central Mexico (^{González et al., 2008}; ^{Torres et al., 2011}, 2017; ^{Quiroz et al., 2017}).

The males with the highest grain yield were 7, 38, 41, 35, 34, 33, 9, 24, 36, 30, 23, 32, 19, 10, 11 (6 362 to 6 994 kg ha^-1) only 7 and 41 were later. There were no significant differences in their plant heights, but those of the cob; they also presented good aspect of plant and cob, similar percentages of lodging, tillering and the most prolific were 38, 11, 10, 41, 9, 35, 33, 34 and 36 (Table 6). This upper fraction could be used in this region of Mexico to generate new hybrids of three and four lines with double purpose: production of more grain and green or dry matter.

Conclusions

The best locations for the evaluation of the genetic material were Atlacomulco, Almoloya of Juárez and Toluca. The interaction between localities and the rest of the interactions was significant in most of the evaluated variables, so the behavior of the genetic material through the evaluation environments is different. It is suggested to give greater importance to specific adaptation.

The female one (CML246 x CML242) contributed to the formation of mestizos of larger dimensions in grain yield, plant and cob heights, tillering and plant and cob aspects; also, had lower averages in male and female blooms, acames of root, stem and total, cob cover, cob rot and prolificacy.

The upper fraction of males recommended for the formation of new double or triple cross hybrids are those identified as 7, 38, 41, 35, 34, 33, 9, 24, 36, 30, 23, 32, 19, 10, 11, 31, 22, 27, 25, 29, 21, 37, 26, 28, 12 and 14.