Introduction

In Mexico it is necessary to increase the production of yellow grain maize (Zea mays L.) to satisfy the demand and reduce the importation of 10 million Mg of whole and cracked grain, for the elaboration of livestock feed, extraction of starches, cereal and snack industry, and other industrial uses. Annually 8.5 million ha of maize are cultivated in Mexico, with production of 22.5 million Mg and an average of 2.8 Mg ha^-1 (^{Turrent, 1994}; ^{Ortiz et al., 2007}; ^{Turrent, 2009}).

Of the national cultivated surface, 1.5 million ha are found at altitudes of 2200 to 2600 m in the High Valleys of the Central Plateau; of these 800 thousand ha are cultivated under rainfed conditions, with late rains that limit the planting date, crop productivity and its exposure to the incidence of early frosts. Under these conditions, in the State of Mexico 300 thousand ha are planted, with a production of 1.2 Mg ha^-1 (^{Ávila et al., 2009}; ^{Espinosa et al., 2010 a}).

The international price of maize increased when this food resource was incorporated to the elaboration of ethanol (^{Ortiz et al., 2007}). Therefore, the use of improved seed is a key element for achieving competitive levels in production, as the environmental conditions have less effect on the improved cultivars ex profeso for a given region and the use of inputs required by the production process are optimized. Improved seed supplies as much as 60 % of the final yield, thus it is a fundamental input (^{Ortiz et al., 2007}; ^{González et al., 2008}).

The improved varieties of yellow grain maize with short cycle, adapted to the agroclimatic conditions of the High Valleys are V-26A, Amarillo Zanahoria, V-31 A, V-34 A, V-35 A, but in this region the highest production of yellow maize is sustained with native varieties. The improved varieties are few; one is V-26A (Cuapiaxtla), which was released in 1980, and is currently in disuse due to the lack of seed production. Another is Amarillo Zanahoria, released in 1990, and is currently in disuse due to lodging and low yield (^{Espinosa et al., 2010 a}; ^{Espinosa et al., 2011}).

To attend the need of yellow grain maize varieties, the Cuautitlán Graduate Studies Department Cuautitlán of the National Autonomous University of Mexico (FESC-UNAM) has generated early cycle varieties (^{Tadeo and Espinosa, 2004}; ^{Tadeo et al., 2004}; ^{Tadeo et al., 2012}). In the National Institute of Research of Forestry, Agriculture, and Livestock (Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias), Experimental Field Valle de México (CEVAMEX-INIFAP) the yellow grain and early cycle varieties ‘V-53A’, ‘V-54-A’ and ‘V-55A’ were generated, of which ‘V-54-A’ and ‘V-55A’ have Breeder’s Titles and are registered in the National Catalogue of Plant Varieties (^{Espinosa et al., 2009}; ^{Espinosa et al., 2010 a}; ^{Espinosa et al., 2011}).

The yellow grain varieties released by CEVAMEX-INIFAP show acceptable yields in delayed crops, tolerance to lodging and advantages under medium productivity conditions with respect to the controls. For the purpose of generating varieties with higher yield, intervarietal crosses were made between ‘V-53A’, ‘V-54-A’ and ‘V-55A’ and other varieties, with simple crosses and lines of contrasting germplasm which include subtropical sources of the High Valleys of Jalisco (^{Ramírez et al., 2013}; ^{Ledesma et al., 2015}), and include lines and simple crosses unrelated to the reference varieties and from diverse ecological regions (^{Ramírez et al., 2013}; ^{Ledesma et al., 2015}).

The non-conventional hybrids are defined as hybrids obtained from the combination of parents that do not comply with the classical conformation of simple, three way or double cross hybrids, integrated with the participation of two, three and four parent lines, respectively. These non-conventional hybrids result from the combination of a variety x variety, variety x hybrid or a variety x line; because of their genetic structure they present the facility for seed production and favorable yields superior to their parents (^{Tadeo et al., 2015}).

The objective of the present study was to determine the productive capacity, heterosis and precocity of 12 non-conventional yellow maize hybrids that utilize as parents improved varieties, lines and simple crosses. The hypothesis was that the combinations of non-conventional hybrids of yellow grain maize improve the yield of their respective parents, maintain precocity and express heterosis in yield.

Materiales and Methods

In the spring-summer cycle of 2012, non-conventional hybrids were formed (^{Espinosa et al., 2013}), for which the improved yellow grain varieties ‘V-53-A’, ‘V-54A’ and ‘V-55A’ were combined, with the line ‘351 #’ of yellow grain maize, which were obtained from CEVAMEX in collaboration with the FESC, and the improved variety 324 # from the Campos Altos de Jalisco Experimental Field (CEAJAL), with good general combinatory aptitude and different germplasm origin (^{Ramírez et al., 2013}). The yellow maize simple crosses ‘CML 460XCML 462’ and CML 461XCML 462’ were also used, which in combination with yellow maize varieties expressed good productive potential (^{Espinosa et al., 2013}).

Furthermore, the varieties ‘V-55A’ and ‘OU2C’ were included, obtained by recombination during two cycles of a balanced compound of 21 S2 lines, of yellow seed, generated from seven simple cross hybrids. Three lines of each hybrid were selected at the start for their precocity, intense yellow color and crystalline texture of the grain. The lines pertain to the conical race and were selected for their yield and precocity in yield assays carried out in FESC-UNAM, in Cuautitlán and by INIFAP. After recombining the balanced compound for two cycles for ‘V-55A’, two cycles of masal selection were applied in CEVAMEX, INIFAP and for ‘OU2C’ in the FESC; in both cases, to gain in precocity at physiological maturity, tolerance to lodging and crystalline texture in the grain (^{Espinosa et al., 2011}).

The combination of crosses: variety x variety, variety x line, variety x simple cross, each parent that comprises them and the origin are included in Table 1. Uniform experiments were established with the cited material, during the spring-summer cycle of 2013 and 2014, and in both years the sowing was carried out in June in: Santa Lucía de Prías, CEVAMEX-INIFAP, in the municipality of Texcoco, State of Mexico (2240 m altitude, climate C(W0)(w)b(I’)g temperate with rains in summer, the driest of the sub-humid climates, with cool and prolonged summers, mean annual temperature between 12 and 18 °C, and annual oscillation of the mean monthly temperatures of 5 to 7 °C (^{García, 2004}); 2) Rancho Almaraz of the FESC, Field 4, of UNAM (19º 41’ 35’’ N and 99° 11’ 42’’ W, 2274 m altitude, climate C (wo) (w) b (i’’) with historic average annual precipitation of 609.2 mm (^{García, 2004}).

Table 1 Origin of non-conventional hybrids (NCH), varieties (V), and simple crosses (HS) of yellow maize grain in the uniform experiments during the spring-summer cycles of 2013 and 2014. 

†INIFAP: Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias; ¶FEC-UNAM: Facultad de Estudios Superiores Cuautitlán campo 4, Universidad Nacional Autónoma de México; §CIMMYT: Centro Internacional del Mejoramiento de Maíz y Trigo ❖

†INIFAP: Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias; ¶FEC-UNAM: Facultad de Estudios Superiores Cuautitlán campo 4, Universidad Nacional Autónoma de México; §CIMMYT: Centro Internacional del Mejoramiento de Maíz y Trigo.

HCN: híbrido no convencional; V: variedad de polinización libre; L: línea.; HS: híbrido simple.

The experimental plot was a row 5 m long by 80 cm wide and population density of 45 000 plants ha^-1, utilized in plantings of June delayed by limited rainfall (^{Espinosa et al., 2010 a}). The experiments were established according to a design of complete randomized blocks, with four replicates and under dry conditions; the sowing was carried out in June of 2013 and 2014 depositing two seeds every 50 cm., covering the seeds with soil. Later, thinning was performed to achieve a population density of 45,000 plants ha^-1; the dose of fertilizer was 80-40-00. In the field, the variables days to male flowering were taken when 50 % of the plants of the plot released pollen, days to female flowering when 50 % of the plants in the plot exposed the stigmas in at least 3 cm, plant height was taken in five plants from the base of the stem to the insertion node of the spike, height of the ear from the base of the stem to the insertion node of the upper ear and expressed in cm.

The harvest was carried out manually in December in 2013 and 2014. In each plot all of the ears were harvested, weighed (field weight) and in a representative sample of five ears the following was determined: percentage of moisture of the grain with an electric moisture determiner (Stenlite), the value was subtracted to 100 and was expressed as percentage of dry matter (% DM); percentage of grain per cob using the quotient of grain weight divided by the grain weight plus cobs (% G); volumetric weight, weight of 200 grains; ear length; rows per ear; grains per row; grains per ear. To obtain grain yield, the following formula was used:

Yield = (PC x % DM x % G x FC)/8600

where

PC: field weight of the total of ears harvested per plot (in kg);

FC: conversion factor to obtain yield per ha, which was obtained by dividing 10000 m² by the size of the useful plot in m² (4 m²); 8600: it is a constant value that makes it possible to estimate yield with a uniform moisture of 14 %, in which the seeds are managed commercially.

The statistical analyses were made in SAS v 9.0 (SAS Institute Inc., 1996). The combined analysis of the two locations and two years of testing considered as sources of variation environment, genotypes and interaction environment x genotypes. The averages were compared with the Tukey test (p≤0.05). The level of heterosis of the hybrid combinations with respect to the mean parent was estimated with the following formulas (^{Márquez, 1988}):

P=(P₁ + P₂)/2

where

P₁ and P₂ are the genotypic values of the parents:

H(F1/P) = F1 - P

where

F1 is the genotypic value of the hybrid.

Heterosis based on the best parent was calculated with the following formula:

H = F1 - Best parent / Best parent.

Results and Discussion

The combined analysis of variance detected statistical differences (p≤0.01) among environments and genotypes in all of the variables (Table 2). The interaction environment x genotype was significant for most of the variables and grain yield, except for plant and ear height and weight of 200 grains. The coefficients of variation were lower than 16 %, which confirmed the high quality and the control of the experimental variability. The average value of general experimental yield (6458 kg ha^-1) showed that production was good, in the delayed plantings and higher than the national average (2.8 Mg ha^-1) (^{Turrent, 1994}; ^{Ortiz et al., 2007}; ^{Turrent, 2009}) and of the State of Mexico (1.2 Mg ha^-1) (^{Ávila et al., 2009}; ^{Espinosa et al., 2010a}).

Table 2 Means squares of the combined analysis for yield and other variables in non-conventional maize hybrids and their parents in two locations of the High Valleys in the spring-summer cycles of 2014 and 2014. 

*p≤0.05. **p≤0.01. REND: yield; FM male flowering; FF: female flowering; AP: plant height; AM: ear height; PHEC: weight per hectoliter; 200G: weight of 200 grains; %MS: percentage of dry matter; %GR: percentage of grain.

Between locations (Table 3) most of the significant differences of the yield behavior (Figure 1) and the other variables were due to the differential of precipitation between the years and some meteorological incidents, such as hail in CEVAMEX in 2014 which stimulated flowering and events of rainfall distribution during the crop cycle and low temperatures in the grain fill stage. Thus, for example, in the Almaraz meteorological station, of the FESC, in June of the 2014 cycle, rainfall of 211.9 mm was measured along with maximum temperature of 24.1 °C; in contrast, in June of the 2013 cycle, rainfall was 117 mm and the maximum temperature 25.3 °C. The availability of moisture in the soil during germination was higher in 2014 and affected yield (Figure 2).

Table 3 Behavior of means among environments of the combined analysis of locations during the years 2013 and 2014. 

AMB: environment; Rend: yield; FF: female flowering; AP: plant height; AM: ear height; %GR: percentage of grain.

Figure 1 Yield of the non-conventional hybrid V-53 A x 324 # and its parents through four evaluation environments. Spring-summer cycle 2013-2014. 

The increment of rainfall and temperature in August-September, according to the data of the Almaraz meteorological station, favored the synchrony of male and female flowering. In contrast, in 2014, lack of moisture in flowering caused 5 d of asynchrony between male and female flowering; this reduced grain yield in this location. The lack of moisture before and during flowering also affected plant height (Table 3). The significant behavior of the percentage of rain coincided with grain yield; that is, the most productive locations were associated with higher percentages of grain. This coincided with what was reported by ^{Ramírez et al. (2013)}, ^{Espinosa et al. (2013)} and ^{Tadeo et al. (2015)}.

The hybrids ‘V-53A x 324 #’ (8185 kg ha^-1) and ‘V-55A x 324 #’ (8080 kg ha^-1) showed yields statistically similar to their male parent in both cases, ‘324 # (9085 kg ha^-1) maintained the precocity of its female parent varieties, ‘V-53A and ‘V-55A’, identified as the most precocious in the High Valleys (^{Espinosa et al., 2010 a}; ^{Espinosa et al., 2011}), but with significant yield superior to each one of them. In the first case, the non-conventional hybrid ‘V-53A x 324 #’ was 25.4 % higher with respect to ‘V-53A’, and 32.6 % in the second ‘V-55A x 324 # with respect to the variety ‘V-55A’ (Table 4). The achievement of the combination of precocity and high yield in these two non-conventional hybrids represents per se an advantage for their adoption in late plantings (Espinosa et al., 2010 a).

Table 4 Mean behavior of nonconventional hybrids and their parents of yellow grain maize in four evaluation environments during the spring-summer cycles of 2013 and 2014. 

*p≤0.05. **p≤0.01. Rend: yield; FM male flowering; FF: female flowering; AP: plant height; AM: ear height; OP2D: Oro Plus 2D; OU3C: Oro Ultra 3C.

The yield of the non-conventional hybrid F₁ (‘V-53A x 324 #’) was outstanding in the 2013 FESC environment; in contrast, for the non-conventional hybrid F₁ (‘V-55A x 324 #’) the best environment was FESC 2013 and was similar to CEVAMEX 2014 (Figure 1 and 2). The yield values for the non-conventional hybrid F₁ (‘V-54A x 324 #’) and its parents showed that this genotype in CEVAMEX 2014 obtained the highest values (Table 4, Figure 3). The combination ‘V-55A x 351 #’ and their parents in the FESC 2013 environment showed the highest yields, and in all of the cases it was observed that the line ‘351 #’ obtained low individual yields (Figure 4). The expressions of yield of the materials is related to the genetic constitution and type of material, in line 351# it propitiated low yields; in contrast, in previous studies V-53A, V-54 A and V-55 A, the latter two released commercially, its genetic constitution exhibited a favorable response to different conditions (^{Espinosa et al., 2010a} and ²⁰¹¹).

Figure 2 Yield of the non-conventional hybrid V-55 A x 324 # and its parents through four evaluation environments. Spring-summer cycle 2013 and 2014. 

Figure 3 Yield of the non-conventional hybrid V-54 A x 324 # and its parents in four environments. Spring-summer cycle 2013-2014. 

In the crosses with highest yield, ‘V-53 A x 324 #’, ‘V-55 A x 324 #’ and ‘V-54 A x 324 #’, the heterosis, based on the mean parent, was 5.0 %, 6.5 % and -0.8 %, respectively, which scarcely stood out in each cross from the high yield of the variety ‘324 #’, which was also the best parent in the three cases. The values of heterosis with respect to the best parent were -9.9 %, -14.9 % and -11.1 %, respectively (Table 5). The above indicated that there was no heterosis in the non-conventional hybrids with respect to ‘324 #’; in the case of this parent, its genetic constitution and being three or four days later than the non-conventional varieties and hybrids, is not favorable in the late plantings, for which it would be recommended. Therefore, the commercial exploitation could be concentrated with greater advantages in the non-conventional hybrids for their precocity. The absence of endogamic depression in the parents of the above three hybrids could also explain the absence of heterosis; although its presence is possible due to the distance from the origin of parents (^{Ramírez et al., 2013}), this did not occur in these three crosses.

Figure 4 Yield of the non-conventional hybrid V-55 A x 351 # and its parents in four environments. Spring-summer cycle 2013-2014. 

Two of the crosses, ‘(CML 461 x CML 462) x V-55A’ and ‘(CML 461 x CML 462) x OU2C’, displayed heterosis in relation to the mean parent of 49.8 % and 56.5 %. In both cases the heterosis could be attributed to the different origin of the parents which do not have endogamy, given that in both cases they are varieties of free pollination. The cross that showed the highest expression of heterosis with respect to the best parent was ‘(CML460 x CML462) x V-53A’ (Table 5); however, the exhibition of heterosis did not place these combinations in the higher levels with respect to the best non-conventional hybrids (^{Tadeo et al., 2012} and ²⁰¹⁵). Here, with the absence of endogamy in the parents, the hypothesis of the divergent origin of the materials is strengthened. The lack of heterosis among the varieties ‘V-53A’, ‘V-54A’ and ‘V-55-A’ (although they come from different programs) and from them with the variety ‘324 #’, indicates that probably despite their different geographic origin, they do not differ in their germplasm source. The above supports the alternative of production of seeds resulting from the cross among non-conventional hybrids and improved populations (simple cross, variety, hybrid varietal), whose origin is different (^{Ramírez et al., 2013}).

Table 5 Average values of heterosis in the yield of yellow grain maizes, with respect to the general average and the best parent in non-conventional hybrids 

Conclusions

Given that hybrids ‘V-53A x 324 #’, ‘V-55A x 324 #’ and ‘V-54A x 324 #’ significantly surpassed in yield the commercially used varieties ‘V-54A and ‘V-55A’, the hypothesis that the non-conventional hybrids would surpass their parents, maintaining precocity was confirmed, therefore, these three non-conventional hybrids would offer advantages for their commercial use in the High Valleys of Mexico.

There was no heterosis among the varieties ‘V-53A’, ‘V-54A’ and ‘324 #’, probably due to their origin, and to the fact that these varieties do not differ in their germplasm source, in contrast to the cross (‘CML 460 x CML 462’) x V-53A’. Therefore, the alternative of seed production with non-conventional hybrids is supported, with improved varieties (simple cross, variety, varietal hybrid), whose origin is divergent.