Introduction

Corn is a grass of economic importance and wide genetic and morphological variability; it is produced in all agroecological regions of Mexico. This favors that this crop is attacked by a wide diversity of pest insects in tropical and subtropical regions, such as northeastern Mexico, where Spodoptera frugiperda (J. E. Smith) or fall armyworm stand out (^{Fernández et al., 2013}; ^{Yáñez et al., 2019}). This insect, in the larval stage, mainly consumes developing leaf tissue, which decreases the photosynthetic area of the plant and consequently the grain yield (^{Neves et al., 2020}; ^{Obok et al., 2021}).

Due to the above, synthetic insecticides are mainly used for its control (^{Burtet et al., 2017}), which can lead to resistance in the insect, elimination of non-target species (^{Neves et al., 2020}) and environmental pollution (^{Guo et al., 2020}). An alternative to reduce these risks is the use of cultivars tolerant to leaf damage caused by S. frugiperda (^{Smith and Clement, 2011}). To obtain these cultivars through genetic improvement, it is necessary to identify and have base germplasm with adaptation to specific agroecological conditions (^{Fuentes et al., 1998}).

Corn-growing agricultural regions in Tamaulipas, as in much of the world, due to climate change (^{Yujie et al., 2021}), present humidity restriction, extreme temperatures and a period of dog days, where the temperature exceeds 40 °C (^{Castro et al., 2013}). This causes a decrease in the growth cycle in the crop (^{Luna et al., 2005}), generating the need to use cultivars with high earliness (^{Pecina-Martínez et al., 2009}), between 50 and 70 days to flowering, spring-summer (SS) and autumn-winter (AW), respectively.

Earliness in corn is mainly estimated by the days from sowing to male and female flowering (^{García et al., 2020}). This is related to high photosynthetic efficiency, adequate distribution of assimilates, lower plant height (^{Dos Santos et al., 2019}) and high harvest index (^{Coutiño-Estrada et al., 2014}). Therefore, within a genetic improvement program for the tolerance of corn to S. fruigperda damage in the dry tropics, it is desirable to use earliness as a complementary criterion for selection and evaluation (^{Preciado et al., 2002}).

However, it is considered that early cultivars have lower yields than late ones, therefore, it is necessary to characterize, in the base germplasm, the specific relationships between earliness and grain yield, and to know the genetic effects that control the expression of variation in earliness (^{Dzib-Aguilar et al., 2011}). The native populations of central and southern Tamaulipas, in addition to wide variability, have potential for high yield, earliness and adaptation to restrictive environmental conditions (^{Pecina et al., 2011}), since they have developed in the environments described above (^{González-Martínez et al., 2020}).

On the other hand, to comprehensively know the gene action that controls the variation of the base germplasm, a phenotypic and genotypic evaluation must be carried out, using characteristics that favor adaptation to the specific agroecological regions of interest (^{Ertiro et al., 2017}). In this sense, there are tools that help to know the genetic components that control the phenotypic variation of corn populations, which allows selecting the most appropriate improvement method (^{González et al., 2016}; ^{Ferreira et al., 2020}).

The diallel designs are defined as a set of simple crosses between a group of n parents (^{Sánchez-Ramírez et al., 2017}), they allow estimating the effects of general (GCA) and specific (SCA) combining ability, reciprocal, maternal and non-maternal effects (^{Griffing, 1956}; ^{Ramírez-Díaz et al., 2019}). The relative proportion of the sum of squares of reciprocal effects, GCA and SCA in relation to the sum of squares of crosses indicates the proportion of influence of each type of gene action on the variation of the germplasm evaluated, where GCA is mainly associated with additive effects, while non-additive effects are associated with SCA and reciprocal effects, the latter are the result of maternal and/or non-maternal effects (^{Hernández and De León, 2021}).

The expression of variation in the days to flowering of corn depends mainly on the effects of GCA and SCA (^{De la Cruz et al., 2003}; ^{García et al., 2020}), there is also influence of reciprocal effects (^{Cervantes-Ortiz et al., 2018}) but there is little or no information on their proportion, due to maternal and non-maternal effects (^{Issa et al., 2018})

It is recognized that the level of genetic diversity of parents influences the expression of reciprocal effects (^{de la Cruz et al., 2003}; ^{Sánchez-Ramírez et al., 2017}), so it is necessary to establish this type of parameters in a specific way, for each population or group of populations to be used as base germplasm for improvement. Based on the above, the objective of this study was to determine the gene action that controls the variation of the earliness of native corn populations of Tamaulipas and the relationship of earliness with grain yield.

Materials and methods

Results and discussion

The statistical analysis showed significance (p< 0.05) between environments and between cultivars for male and female flowering; on the contrary, there was no significant interaction (p≥ 0.05) for environment × cultivar (Table 2). The above demonstrates that the differential environmental conditions determined by the combination of agricultural cycle and management strategy of S. frugiperda had a similar effect for the expression of days to flowering of each of the cultivars evaluated; that is, the cultivars presented a similar response to the environmental changes determined by the combination of agricultural cycle and management strategy of S. frugiperda.

No significant effects of GCA were observed for the variables evaluated and there was no statistical significance for the interaction of GCA × environment (Table 2). On the contrary, there were significant effects of SCA, reciprocal and maternal for both variables evaluated, the significant effect of SCA indicates that within the evaluated germplasm there is variability for the gene expression of male and female flowering due to non-additive effects (^{García et al., 2020}).

The significance for reciprocal effects showed that the average number of days to male and female flowering of the direct cross (i, j) is different from that of the reverse cross (j, i) for at least one pair of parents (i, j), so populations that serve as female or male parents should be specifically considered (^{Núñez-Terrones et al., 2019}). This was the result of significant maternal effects, which indicates that female and male flowering expression in these crosses was controlled by cytoplasmic genes (^{Mahgoub, 2011}).

The expression of variation among the evaluated cultivars is explained in 52.91% for male flowering and 44.75% for female flowering by SCA effects (Table 2), while 38.22% of the variation in male flowering between cultivars and 42.74% for female flowering was due to reciprocal effects (Table 2). This indicates that non-additive effects were more important in the expression of genetic variation, both for male and female flowering; that is, the observed number of days elapsed from sowing to flowering in crosses is mostly due to dominance or non-additive effects and this may be due to genetic divergence between populations for these variables (^{De la Cruz et al., 2003}).

Therefore, the improvement of earliness in these populations should be done through processes of recurrent reciprocal selection or hybridization (^{Rodríguez et al., 2020}) and the formation of a particular cross should specifically consider the lines that will serve as a female or male parent (^{Núñez-Terrones et al., 2019}). For male flowering, crosses TML₃S₃×VHA and Cam×VHA presented significant SCA with a negative estimator (Table 3), which indicates that they had fewer days to male flowering than expected according to the GCA of the populations that participated as parents (^{Mahgoub, 2011}), with an average of 63.4 and 63.7 days respectively (Table 4).

Similarly, for the days to female flowering the cross TGL₂S₃×PWL₆S₃ presented significant SCA with a negative estimator (Table 3) with an average of 65 days (Table 4). Taking into account the above and considering the absence of effects of GCA in this germplasm (Table 2), it can be considered that this behavior is due to a low accumulation of additive genes for male and female flowering in these crosses, with predominance of the parents with expression of over dominance for these variables, where the existence of allelic interaction (SCA) in the evaluated progeny is demonstrated (^{Gutiérrez et al., 2002}).

Thus, it was observed that crosses with significant SCA had fewer days to male and female flowering compared to their parents (Table 4). This is consistent with what was established by ^{de la Cruz et al. (2003)}, who point out that, in general, the development of crosses is faster than that of their parents; mainly due to the hybrid vigor or heterosis, which favors high rates of development and growth compared to their parents. The differences shown for male and female flowering between the crosses evaluated may be indicative of genetic diversity among the parent populations, as noted by ^{Guillén-De la Cruz et al. (2009)}.

For male flowering the crosses PWL₁S₃×VHA, TML₃S₃×PWL₆S₃ and Cam×VHA showed significant reciprocal effects and for female flowering PWL₁S₃×TGL₂S₃, TGL₂S₃×PWL₆S₃, TGL₂S₃×VHA and TML₃S₃×PWL₆S₃ (Table 5). The above means that direct cross is different from reciprocal cross (^{Picón-Rico et al., 2018}) and it is verified in Table 4, where it can be seen that, between direct and reciprocal crosses, there are differences of 3.2, 3 and 2.8 days respectively, for male flowering and for female flowering of 2, 2, 2.5 and 2.3 days (Table 4).

Considering all the above, for male and female flowering, VHA×Cam stands out with 62.4 and 65.2 days respectively, values lower than the general mean minus the standard deviation (Table 4). Demonstrating a greater earliness, a favorable characteristic for its adaptation to short growth cycles (^{Luna et al., 2005}), such as those that occur in most agroecological regions of Tamaulipas (^{Pecina-Martínez et al., 2009}).

Significant maternal effects were determined in the parents PWL₆S₃ and VHA, the first with a positive estimator and the second with a negative estimator; both for male and female flowering (Table 6). The above indicates that some of the crosses that had VHA as their maternal parent presented a lower number of days to flowering compared to their reciprocal crosses; as observed in PWL₁S₃×VHA and Cam×VHA with 66.3 and 64.9 days to male flowering respectively, while VHA×PWL₁S₃ and VHA×Cam recorded 63.1 and 62.4 days respectively (Table 4 and 5).

Conversely, when PWL₆S₃ was the maternal parent, at least one cross had a higher number of days to male and female flowering compared to the cross where this population participated as the paternal parent. An example of this is PWL₆S₃×TML₃S₃ with 65.4 and 67.5 days to male and female flowering respectively, compared to TML₃S₃×PWL₆S₃ with 62.5 and 65.2 days (Table 4). The above shows that the variation of the expression of male and female flowering is controlled by cytoplasmic or extranuclear inheritance (^{Sánchez-Ramírez et al., 2017}), which must be taken into consideration when defining the order of the parents in the crosses to be carried out (^{Andrio-Enríquez et al., 2015}).

When the four environments evaluated were considered, no significant correlation (p> 0.05) was observed between grain yield per plant and days to male and female flowering (Table 7). This same result was observed in the AW agricultural cycle, in both management strategies of S. frugiperda and in the SS cycle only with application of emamectin benzoate.

Conversely, in the SS agricultural cycle when no insecticide was applied to S. frugiperda, a significant and positive correlation of grain yield per plant with male and female flowering was observed (Table 7), this indicates that the earliest cultivars had a higher grain yield per plant. This is due to the fact that the SS cycle presents environmental conditions of higher temperature at the beginning of the growth cycle compared to the AW agricultural cycle (^{González et al., 2016}; ^{Reyes, 2017}a), which reduces the length of the growth period (^{Ruiz-Corral et al., 2002}; ^{Kalimuthu et al., 2015}), favoring the development and growth of early cultivars (^{Luna et al., 2005}).

In this agricultural cycle, a higher incidence of S. frugiperda has been reported, compared to the AW cycle (^{Cantú et al., 2010}; ^{Peterson et al., 2017}) and in this study when no insecticide was applied, there was a greater average leaf damage caused by S. frugiperda, 2.96 on a scale of 0 to 5, compared to the AW agricultural cycle without insecticide application of 1.81 and with insecticide application in both cycles without apparent leaf damage (data not shown), which allows inferring that cultivars of greater earliness have better adaptation to environmental conditions with the presence of pests and high temperature.

Conclusions

The variability for male and female flowering between cultivars is mainly explained by specific combining ability and reciprocal effects; the latter are the result of maternal effects. The earliness observed in the crosses TML₃S₃×VHA, Cam×VHA and TGL₂S₃×PWL₆S₃ shows cytoplasmic inheritance. In restrictive environments of high temperature and presence of S. frugiperda, early cultivars have higher grain yields.