Introduction

The Lagunera region is the main region of cattle bovine milk production in the whole country (^{Ramírez et al., 2016}), therefore, it is the area that requires the most forage. In the region, there are more than 400 thousand dairy cows, which produce approximately 925 000 tonnes of dry manure per year, with an average of 1.6% nitrogen (N) (^{Magri and Teira-Esmatges, 2015}; ^{Ramírez et al., 2016}), which provides 14 800 tons of nitrogen (N) per year. The use of these organic waste is increasingly important as rational recycling of nutrients, helps plant growth and return to the soil the elements extracted during the production process (^{Zhao et al., 2014}).

Among the forages that are most consumed and produced in the region is corn, it is the second important crop in the Comarca Lagunera after alfalfa, with 32 491 ha (^{SIAP, 2012}) and occupies approximately 36% of the surface planted at the national level (^{SIAP, 2014}). On the other hand, the constant growth of the number of head of cattle in the region, demand the increase in the production of fodder maize and raises the need to identify materials with good forage characteristics, so among the selection criteria to identify suitable hybrids, is that which presents high grain yield, productivity and high forage quality (^{Sánchez et al., 2011}).

In addition, livestock activity generates large amounts of manure with nutritional potential, which can be used to make a sustainable production system, however, it is important to carry out local studies on N mineralization of manure, to estimate the manure doses of manure precise way (^{Flores et al., 2009}; ^{Azeez and Van Averbeke, 2010}). Therefore, the objective of this work was to evaluate the forage production on dry basis and the protein content of three fodder corn hybrids, using different amounts of vermicompost and bovine manure for fertilization, as well as to evaluate the effect of these treatments in the physical-chemical properties of the soil.

Materials and methods

Results and discussion

The analysis of variance showed significant differences (Table 1, p≤ 0.05) in the treatments (Trat), for the variables: green fodder (FV), dry fodder (FS), plant height (AP), cob width (AM)), length of cob (LM) and number of rows per cob (NHM), which suggests different effects of the treatments. Highly significant differences (p≤ 0.01) were also observed for stem diameter (DT). In the hybrid variation factor (Hib), highly significant differences (p≤ 0.01) were observed for the variables plant height (AP), stem diameter (DT), number of internodes (NE), length of cob (LM) and number of rows per corn (NHM); and significant differences (p≤ 0.05) in the variables of green forage (FV), cob width (AM) and crude protein (PC). The significant differences observed in the plant variables related to the production of forage (Table 1) show the effect of the treatments as well as the genetic differences between the evaluated hybrids, noting that organic fertilization is an alternative of agricultural production at the same time that a sustainable use of manure is made (^{Annicchiarico et al., 2011}).

Regarding the soil variables of the initial sampling (Table 2), the organic matter (MO1) and electrical conductivity (CE1, CE2) showed highly significant differences (p≤ 0.01) in the treatment variation factor (Trat), while the organic matter (MO2) and pH (pH1 and pH2) showed significant differences (p≤ 0.05). No significant differences were observed for nitrates.

In the analysis of variance of the final physical-chemical variables of soil (Table 3) significant differences (p≤ 0.05) were observed for the treatments in the variables MO3, pH3, pH4 and highly significant differences (p≤ 0.01) for CE3, CE4, NO₃3, NO₃4.

In the comparison of treatment means for the forage variables (Table 4), green tilapia stood out, treatments T3 and T6, with 70 t ha^-1 and 68 t ha^-1, respectively. In stem diameter (DT) the best treatments were T2, T3, T5 and T6, with a value between 2.6-2.7 cm. The T3 treatment exceeded the others in plant height (AP), with a value of 2.9 m; in cob width (AM), treatments T2, T6 and T3 stood out, with values 5.2, 5.1 and 5.1 cm, respectively; in the length of cob (LM), the best treatments were T2, T3, T1 and T6, with 19.6, 19.5, 19.3 and 19.1 cm, respectively; the highest number of rows per cob (NHM) was observed in the T6 treatment with a value of 15.6 cm.

The observed values for green forage were higher than the regional average of 47.8 t ha^-1 (^{SIAP, 2014}), and although there were no significant differences in dry forage, all treatments exceeded that reported by ^{Salazar et al. (2007)} who obtained 23.6 t ha^-1 with 160 t ha^-1 of cattle manure. In the crude protein (PC) variable there were no statistically significant differences between treatments, and the values obtained were similar to the highest values reported by ^{Tadeo-Robledo et al. (2012)} but lower than the national average reported by ^{Nuñez (2006)} of 10.3%. The decrease in crude protein content could be the result of the loss of leaf biomass coupled with a higher proportion of stem in the total biomass, which are deficient in crude protein content (^{Tariq et al., 2011}).

Regarding the comparison of means of the physico-chemical variables at the beginning of the experiment (Table 5), in the organic matter at the two depths the best contents were observed in the treatment T6 and T3, and the lowest values in the control T4 which is corresponding to the amounts of manure applied. Regarding pH, no significant differences were observed and the average value was 7.7. Regarding the CE, the highest values were observed in the treatments T3 and T6, and the lowest in the treatment T4, which are consistent with the applied doses of manure and vermicompost in each treatment and are a reflection of the addition of nutrients and salts in applied organic matter (^{Carmo et al., 2016}). Regarding nitrates, they were considered statistically equal in all treatments in the initial sampling.

In relation to the comparison of means of the physico-chemical variables at the end of the experiment (Table 6), only in the organic matter and in the depth of 0-15, the mean values presented significant differences being the treatments T3 and T6 where they observed the highest values. In relation to the pH there was a slight increase in the two depths with respect to the initial values, being the highest value of 7.9. In relation to the CE, a decrease of the salts with respect to the initial content was observed, and the highest value was 4.2 dS m^-1 in the depth of 15 -30, observing the highest values in the treatments with the highest dose of manure (T3 and T6), as well as a slight increase in CE in the treatments that had vermicompost, but within the range tolerated by the corn crop (^{Khorasani et al., 2012}). Likewise, a tendency of increase in the amount of nitrates was observed, as the manure dose increased in the treatments, with the best treatments being T3 and T6.

All the treatments that took manure and vermicompost (except T4) generated changes in the physical-chemical properties of the soil. The T6 treatment initially gave a value of 4.2%, while the T3 treatment was better in the depth of 15-30 (MO2) with a value of 3.2%, being these two strata (0-15, 15-30 cm) the most important for root exploration and where most of the absorption of nutrients occurs (^{Andrades and Martínez, 2014}). For the variable electrical conductivity (CE1, CE2, CE3, CE4) all the treatments generated changes in the amount of soluble salts in the soil, and the highest values were observed in the initial sampling, which could have damaged the absorption of nutrients in solution (^{Carmo et al., 2016}).

In the variable nitrates the values were higher in the initial sampling in the depth 15-30 cm (NO32), highlighting the T6 treatment with 32.1 ppm, a value lower than that reported by ^{Salazar et al. (2007)}, who report a value of 70 ppm for the same treatment and at the same depth; in the depth of 0-15 cm (NO₃1) the T3 treatment stood out with a value of 25.4 ppm, also lower than that reported by ^{Salazar et al. (2007)}, for the same treatment and depth. This could be due to factors such as moisture content, microbial activity, temperature and the state of the manure, which accelerate or delay the availability of nutrients (^{Trejo et al., 2013}).

In relation to the agronomic behavior of the hybrids (Table 7), green hybrid forage was better HT-6806Y with 69.9 t ha^-1, in plant height, the hybrid DAS-2306, with 3.14 m, and in diameter Stem (DT), the hybrid RX-717 stood out with a value of 2.75 cm, this being a variable directly related to the production of green forage, as mentioned by ^{Wong-Romero et al. (2006)}, who indicate that the weight of corn with totomoxtle and the weight of stems with spikes, are the most important for green forage yield.

In the raw protein variable, hybrids RX-717 and HT-6806Y stood out, with 9.1 and 8.9% respectively, which are within the range (7.5-10.3%) reported by ^{Jurado et al. (2014)} for commercial hybrids, and those who mention that since the production of dry matter and forage quality of maize have a direct effect on the milk ha^-1 production potential, these factors should be taken into account in the selection of hybrids of corn for forage production. The percentage of crude protein is an important variable to determine the quality of the forage (^{Lauer et al., 2001}; ^{Saha et al., 2017}) and this content depends on the genetic potential of the hybrid or variety, and agronomic management (^{Silva et al., 2005}).

The variables width of cob (AM), cob length (LM), and number of rows per cob (NHM) are characteristics of maize hybrids with high yield and quality of forage and are components of the characteristic called cob percentage, considered as one of the quality attributes of fodder corn as it contributes to the net energy of lactation (^{Núñez et al., 2010}). In the variable width of cob (AM), the hybrid RX-717 stood out, as in the variable length of cob (LM) and number of rows per cob (NHM), which makes this hybrid a good candidate to be followed exploiting commercially, since although there were no statistical differences in dry forage, it was the one with the highest production; In addition, the values observed in dry forage in this work exceeded the values mentioned by ^{Jurado et al. (2014)} for the state of Chihuahua.

On the other hand, the identification of outstanding genotypes with high nutritional value of forage is an important aspect in terms of feeding requirements of livestock (^{Lauer et al., 2001}), in this case the best hybrid was RX-717 since it excelled in dry forage, crude protein, plant height, stem diameter, width and length of cob, and number of rows per cob.

The correlations between plant variables (Table 8) show positive and highly significant correlation, between raw protein with cob length; This correlation is important since the cob stores a large amount of protein and energy that contributes to improve the quality of the forage (^{Sánchez et al., 2013}) and it was in the hybrid RX-717 that the highest values of cob length and crude protein were observed similarly, the crude protein variable correlated positively with the number of internodes, which means more leaves and therefore more protein, since the leaves provide important values of protein to the forage (^{Hassan, 2011}). The variable of green forage strongly correlated with the variable length of cob, there was also a positive and highly significant correlation between cob width and stem diameter.

Although none of the hybrids used showed remarkable superiority in the production of green matter and dry matter, they exceeded the national average yield (47.61 t ha^-1) in green fodder under irrigation (^{SIAP, 2014}) and considering that they stood out significantly in some variables that are directly related to production and quality, they can be recommended for future experiments. In addition, the hybrid RX717 proved to be better in the production of cobs, both in length and in diameter, which determine the yield and quality of forage.

Conclusions

Manure is a good source of nutrients and can be combined with other sources of nitrogen, to improve plant nutrition and therefore improve some of the variables that directly influence the performance and quality of forages. The T3 treatment with 160 t ha^-1 of solarized manure + 3 t ha^-1 of vermicompost was where the highest production of green fodder, dry fodder and protein content was obtained, surpassing the T6 treatment of only 160 t ha^-1 of solarized manure, so the positive effect of the addition of vermicompost is deduced. Although it is known that soil fertility is a determinant of crop productivity, the genotype of the cultivar used is also determinant to achieve high yields of forage and good quality, as an efficient cultivar in the use of nitrogen can still have acceptable yields with low doses of nitrogen.

In this work no differences were observed in the production of dry forage, but in raw protein content, and in green forage, the yields were very similar among the hybrids evaluated, which denotes the genetic potential of the three hybrids, in particular the Hybrid RX-717 that excelled in forage quality characteristics. The correlations found, indicate that cultivars with high production of green forage tend to have high yields of dry forage, and that high values of crude protein content will be associated with high values of cob length.