Yield

The analysis of variance showed that for the variety factor, there were significant differences (p≤ 0.0001) for all the parameters evaluated. Regarding the FFN, all the parameters were different, except for the harvest index (IC), in contrast to that reported by Barraco et al. (2007), who did not find significant differences in yield when fractionating N-fertilization. Variety*FFN interactions were not significant for dry straw weight (PSP), total dry weight (PST) and IC (Table 2).

The yields and their components obtained in the different varieties are presented in Table 3. With respect to the PSG, the Gema C2004 and Urbina S2007 varieties showed yields similar to those reported by ^{Solis et al. (2007)}, where he mentions that the average yield in The Bajio is 5 t ha^-1, the variety Nana S2007 showed a higher yield in the range of 2.5-3.5 t ha^-1, reported by Villaseñor et al. (2011), this may be due to the fact that this range was determined in rainfed sowing, since this variety shows tolerance to drought. Nana F2007 showed the highest PS; however, 70.3% was straw, which is reflected in an IC of 0.31, in contrast to Gema C2004 who obtained the highest PSG and the highest IC.

Regarding the FFN (Table 4), regardless of the variety, the absolute control obtained the lowest grain yields, with 3116 kg ha^-1, while the best was obtained by fractionating the fertilization at 30-70-00 and 33-33-33, in the order mentioned. When a single application of fertilizer-N is made, the yields are low and do not exceed 4 850 kg ha^-1. Also by not applying fertilization-N at the time of planting, as it is being promoted in the The Bajio region because the demand for N in the initial stages of cultivation is minimal, the yields are lower compared to the fractional application, including the sowing.

The parameters PSG, PSP, PST were lower in the absolute control. The fractionation that showed the highest values of PSG was 30-70-00 with 6822 kg ha-1; however, the IC did not show significant difference. ^{Galantini et al. (2007)}, studied the fertilizer-N application moments and concluded that when there is a greater availability of N, the development of fertile spikes is favored, they also found that the weight of the grains tends to decrease if the whole N is applied at the moment of the sowing and increases when applied in the tillering, which is related to the reduction of losses and the early translocation of assimilated to the grain. On the other hand Waddington et al. (1986); Siddique et al. (1989); Slafer and Andrade (1991), concluded that there is a directly proportional relationship between the number of m-2 grains and yield.

In the interaction variety*FFN, significant statistical difference was observed in the components of yield, for dry weight of grain the fractionation 30-70-00 showed the highest values in each of the varieties, the FFN influenced in the yields in each one of the varieties, the same tendency of the subdivisions was also observed, the best yields were presented when the fertilizer was fractioned in 30-70-00, followed by 33-33-33. Probably due to the reduction of losses and the availability of N in the critical stages of grain formation and filling.

Seed quality

In the analysis of variance, highly significant differences were observed between the varieties and FFN for the variables PH, PMS, GE and LP; in the variety*FFN interaction the variables PH, GE and LP also presented a highly significant difference, for the variable PMS there were no statistical differences (Table 5).

In the results of the comparison of means test for the varieties (Table 6), it was observed that Gema C2002 presents a value of 77.9 kg hL^-1, superior to the rest of the varieties. ^{Vázquez-Mendoza (2013)}, found differences in the PH of commercial varieties of wheat. This parameter is important because the larger it is, the higher the proportion of starch and, consequently, the greater the extraction of flour. Therefore, the hectoliter weight is a good estimate of both the physical quality of the grain, and the milling quality. With respect to the PMS, it was observed that the Urbina S2007 variety presented a value of 47.37 g, followed by Gema C2002 with 44.38 g and Nana F2007 with 39.86 g, these results are consistent with those reported by ^{Solis et al. (2009)}, and is due to genetic variation among varieties.

In the GE test the results were between 90-93%, they are within the standard reported by the ^{SNICS (2013)}, which mentions that a minimum germination percentage for basic seed is 90%. ^{Khabiri et al. (2012)} evaluated 11 wheat genotypes and found a significant difference of this variable and may be due to several factors, among which are the cultivar, genetic purity, environment and maternal nutrition, among others. For the LP recorded on day seven, the varieties Nana F2007 and Urbina S2007 did not show statistical differences with 7.19 and 7.25 cm respectively, the variety Gema had the lowest value with 6.74 cm.

In the comparison of means for the FFN (Table 7), it is observed that in the PH and PMS variables the treatments 00-30-70 and 00-50-50, in which no-N fertilizer was applied to the sowing, showed the higher values, this is because the late application of N increases the protein content of the grain, this effect was reported by ^{Echeverría and Studdert (1998)}, they found that the protein content in the grain responded to the interaction of applications to the planting and time dds. The treatments 00-50-50 and 00-30-70 in which no fertilizer was applied to the seeding presented the highest values in PMS 46.5 and 46.6 g, respectively. With respect to the percentage of GE, the lowest value was shown by the witness, the subdivisions that are within the SNICS standards (2013) are: 50-50-00, 00-50-50, 70-30-00, 30-70-00 and 33-33-33, it was found that there were variations according to the subdivisions of this fertilization.

Regarding the interaction variety*FFN, for the variable PH Gema C2004 showed the highest value (81.17 kg ha^-1) when the fertilizer was divided into 00-30-70, in PMS this same fractionation reached the highest values in the variety Urbina S2007, this is due, as mentioned above, that the low availability of N at the time of planting increases the quality characteristics of the seed; however, as shown in Table 4, the yield with this fractionation was among the lowest. In contrast, the percentage of GE in the three varieties the highest values were found in the fractions 30-70-00, 70-30-00 and 33-33-33.

Vigor of seedlings

When analyzing the characters of seedling vigor (Table 8), highly significant differences were observed in the characters VE, IE, and ITE. Similarly, in the decomposition of the mean squares for the FFN, significant differences were observed in all the characters evaluated. In the variety*FFN interaction, the characters that showed significant differences were VE, %E, IE, ITE, LPt and IV-II. The statistical significance of the FFN indicates that the way to fractionate the N-fertilizer has both positive and negative effects on seed and seedling vigor compared to the absolute control and regional control. ^{Salazar et al. (1996)}; ^{Wuest and Cassman (1992)}, observed that in late applications of N-fertilizer in wheat, the N moves more efficiently to grain, meanwhile, ^{Rivera-Reyes et al. (2009)} report that with higher protein content in oat seed, there is also more phytic acid content, which is reflected in higher physiological quality.

Regarding varieties (Table 9), it can be observed that in VE the highest value was obtained by Nana F2007 with 12.07 Pt d^-1, followed by Gema C2002 and Urbina S2007 with 11.49 and 10.16, respectively; for %E Gema C2002 and Nana F2007 were statistically equal with values of 90.6% and 91.4%, the highest value was 92% observed in Urbina S2007, the largest IE was 5.98 for Urbina S2007, followed by Gema C2002 and Nana F2007 with 5.64 and 5.39 days; in relation to the ITE the varieties Urbina S2007, Gema C2002 and Nana F2007 obtained 0.065, 0.062 and 0.059 and for PSPA the highest value was presented by Gema C2002 with 0.45, between Nana F2007 and Urbina S2007 there were no statistical differences. For the IV-II Gem C2002 showed the highest value with 40.8; similar results were reported by ^{Khabiri et al. (2012)}, when evaluating different genotypes of wheat found significant differences in the speed of germination. Regarding variables IE, E, PSPA, ^{Martínez-Solis et al. (2010)}, observed the same trend among maize genotypes. On the other hand, the parameters LPt and IV-I did not show significant differences, these differ with that reported in the corn crop by ^{Cervantes et al. (2006)}, who found significant differences in all parameters.

In the FFN, the comparison of means indicates that there is a significant difference in all the variables evaluated (Table 10), the best PEs were presented in the divisions 33-33-33, 30-70-00 and 70-30-00, the Seedlings that showed the highest PSPA were those of treatments 00-30-70, 30-70-00, 00-100-00 and 00-50-50. The seedlings that reached greater height were those corresponding to the fractionation 00-30-70 with 13.7 cm. In relation to IV-I and IV-II, the treatment that showed the highest values was 30-70-00.

In the interaction variety*FFN, the comparison of means indicates that for the PE Gema C2004 shows the best percentages in the fractions 30-70-00, 70-30.00 and 33-33-33, like Nana F2007 and Urbina S2007, but these two fractions also obtained statistically equal results in 50-50-00.

For the PSPA values, the best fractions were 00-50-50, 30-70-00 and 00-70-30 for Gema C2004, 30-70-00 and 00-50-50 for Nana F2007, and for Urbina S2007, 00-100-00 and 00-30-70. None of the fractions coincided in the three varieties, this indicates that this variable may be linked more to the genotype than to the FFN, in the variable LPt the same behavior was observed.