Introduction

The problem with ruminants feeding in the tropic is the low quality and availability of grass throughout the year (^{Ramirez et al., 2010}), when cattle consumes only forages, the ingestion of energy could be insufficient for obtaining acceptable production levels (^{Owens et al., 2010}).

The conservation of forages by using silages represents an option to ensure the supply of nutrients and increasing the productivity on bovine and ovine (^{McGeough et al., 2010}); in the silage, sugars in the forage are fermented by anaerobic bacteria, to produce lactic acid (LA) and inhibit the development of other microorganisms (^{Basso et al., 2014}).

In Mexico, the use of trees and fodder species to feed livestock is a common practice in association with gramineae, the quality and nutritious value of the silages of tropical gramineae can be increased when using fodder from trees and bushes, to this effect, Leucaena leucocephala and Brosimum alicastrum stand out among other species (^{Cardenas et al., 2003}); it is necessary to increase the number of research studies on the utilization of these species in gramineae silages of high energetic value like fodder maize.

The objective of this work was to evaluate the effect of including two fodder species into the quality of fodder maize silage in eastern Yucatan, Mexico.

Material and Methods

This work was carried out at the Instituto Tecnologico de Tizimin in the state of Yucatan, Mexico, from June to October, located between 19°40’ north and 87°32’ west. The climate in this region is warm and sub-humid, with a rainy season between June and October, with an annual temperature from 24.5 to 27.5 °C (^{Duch, 1988}).

By means of a completely randomized design, with a factorial adjustment 2 * 3, six treatments plus one control were obtained, in order to asses two fodder tree species (Leucaena leucocephala and Brosimum alicastrum) and three addition levels (15, 30, and 46 %), with four repetitions. Based on the fresh weight, the treatments were the following:

Control (Zea mays).

Z. mays 85 % + Leucaena leucocephala 15 %.

Z. mays 70 % + L. leucocephala 30 %.

Z. mays 55 % + L. leucocephala 45 %.

Z. mays 85 % + Brosimum alicastrum 15 %.

Z. mays 70 % + B. alicastrum 30 %.

Z. mays 55 % + B. alicastrum 45 %.

The harvesting of the whole maize plant (Zea mays) was realized at a flocculent grain state, and the foliage of the fodder plants including stems of ≤1 cm in diameter, petioles and ripe plants of >10 years of age, the forages were ground to obtain a particle size between 0.5 and 1.5 cm. The materials were mixed by treatments and were ensilaged in plastic microsilos with wide opening (1.8 L).

The microsilos were opened on the 60th day, and then the analysis of the chemical composition and fermentative characteristics of the silage was realized. The chemical composition was evaluated by determining the dry matter (DM), by desiccation in stove at 60 °C for 48 hours (^{Pichard et al., 1992}), crude protein (CP), by means of Kjeldahl’s method (^{Galyean, 1980}); acid and neutral detergent fiber (ADF, NDF), by means of the detergent method of determination of cell walls (^{Van Soest et al., 1991}). The fermentative characteristics were determined from the fresh and macerated material, including pH, percentage of lactic acid (LA), ammoniacal nitrogen as percentage of the total nitrogen (N-NH₃/TN), and volatile fatty acids (VFA), in order to correct the content of dry matter (^{Tejada, 1983}).

The variables were analyzed by means of the General Linear Models procedure (^{SAS, 2002}), through a fixed effects model. The initial fixed model included the effect of the fodder species, the inclusion percentage in the silage and its interaction, but because the interaction was not statistically significant, it was excluded from the final analysis. In order to estimate differences within the fixed effects, a comparison analysis of mean proportional was realized by using Duncan’s multiple range test (^{Steel & Torrie, 1980}).

Results and Discussion

The results from the chemical composition of the silages are shown on Table 1. Significant differences (p<0.05) were found to CP and ADF because of the fodders’ addition percentage.

The fermentative characteristics of the silages are shown in Table 2, significant differences were found (p<0.05) in pH, and LA due to the addition percentage, and the content of N-NH₃/TN because of the fodder species and the addition percentage.

The average and standard deviation of dry matter for the silages was 32.2 ± 0.6 %, and there were no significant differences (p>0.05) because of the assessed factors. The content of DM was found within the range reported by ^{Demirel et al. (2006)}, to achieve a state of silage of tropical gramineae (24 to 35 % DM). ^{Sibanda et al. (2007)} did not obtain significant differences (p>0.05) in the content of DM in mixed silages, unlike ^{Apizar et al., (2014)}, who quantified significant increases (p˂0.05) in the content of DM, when siloing sorghum with mulberries (24.8 to 29.6 % DM). An adequate content of DM in mixed silages, ensures homo-fermentative fermentation and the production of lactic acid.

The content of crude protein was 7.7 ± 2.2 %, and significant differences were observed (p<0.05) as a result of the addition levels from fodder trees, the CP increased by 56.5 % in average when including fodders in respect of the control. The effect of adding fodder trees onto tropical silages, has been documented by ^{Castillo et al. (2009)}, who detected a liner tendency on the increase in the content of CP, in response to the addition of Vigna radiata in maize silages; ^{Ojeda & Diaz (1991)} quantified an increase of 5.5 to 11.4 % of CP, when adding 20 % of Lablab purpureus onto the Pannicum maximum silage; ^{Boschini (2003)} and ^{Mora (2010)}, on their part, report an increase in the content of CP from 9.0 to 14.1 % in the maize silage, when adding white mulberry (Morus alba); ^{Phiri et al. (2007)}, observed contents of CP of 13.2 and 14.8 % in mixed maize silages with Acacia boliviana and L. leucocephala respectively. The addition levels of L. leucocephala and B. alicastrum, caused a linear increase in the content of CP of the maize silage.

The average contents of NDF and ADF in the silages were 53.7 ± 1.2 and 36.4 ± 3.1 % respectively, significant differences were found (p<0.05) in ADF, due to the effect of the addition levels of fodder trees. The average values were found within the range reported by ^{Cubero et al. (2010)} for the maize silages (58.6 to 66.5 % and 37.0 to 40.3 %, for NDF and ADF respectively). When including 45 % of fodder trees in the maize silage, the content of ADF increased by 16.9 %, which has been documented by ^{Castillo et al. (2009)}, who reported a significant increase (p˂0.001), in the content of ADF, when increasing the addition of mung bean (Vigna radiata) in maize silages. The content of ADF in ruminant diets, correlates with the indigestible fraction of the material (^{Ayala et al., 2006}), an increase in the percentage of addition of L. leucocephala and B. alicastrum, above 30 % in maize silages, could compromise its quality and nutritional value.

The average pH of the silages was 3.9 ± 0.3, and significant differences were found (p<0.05) because of the addition levels of the fodder trees. The pH from all the silages was found to be under the acceptable maximum value (≤4.3), in order to achieve acidification on tropical silages (^{Ojeda et al., 2006}). The treatment with 45 % of fodder trees, had a 13.2 % of pH higher than the others (p<0.05), which indicates an increase of pH the addition of fodder trees in the silage increases (≥ 455), probably because of the high buffer capacity of the fodder trees, for their greater content of nitrogen and minerals (Ca and P), in comparison to the gramineae. ^{Alpizar et al. (2014)}, report significant increases (p<0.05) of pH, beginning with 75 % of addition of white mulberry (M. alba) in sorghum silages (Sorghum bicolor). Meanwhile, ^{Suarez et al. (2011)}, corroborated that the high values of pH are characteristics of mixed silages (gramineae + fodder trees), nevertheless, its usage is justified by the increase of the proteic fraction and plus, it does not interfere with the activity of homo-fermentative bacteria.

LA is produced by the metabolism of homo-anaerobic and heterolactic bacteria and is the organic compound that affects the most on the acidification of the siloed material (^{Ojeda et al., 1990}). The average content of LA was 4.0 + 1.0 %, and significant differences were found (p<0.05) caused by the effect of addition levels of the fodder trees. The content of acetic acid (AA) and butyric acid (BA) was 0.34 ± 0.13 and 0.07 ± 0.03 % respectively, with no significant differences (p>0.05) because of the assessed factors. All the treatments have a LA content greater than 2 %, considered as minimum in order to achieve the stabilization of tropical silages (^{Esperance et al., 1981}). The contents of AA and BA were found in low levels, in comparison to what was reported for the mixed silages (gramineae ± fodder trees), ^{Titterton et al. (1999)} reportes average values of 1.8 % and 0.6 %, for AA and BA respectively, ^{Cardenas et al. (2003)}, on the other hand, determined values of 3.0 and 0.5 % for AA and BA respectively; which supposes a homolactic-type fermentation (^{Villa et al., 2010}). The mixed silages contained more LA in respect to the control treatment, LA increased on average by 25 and 34 % in silages of L. leucocephala and B. alicastrum respectively, which indicates that the addition of fodder trees does not interfere with the fermentative process of the silage; this effect was corroborated by Tjandraatmadja et al. (2003), who determined high values of LA (>4.1 %) in mixed silages containing 33 % of leguminous plants. According to the addition levels of the fodder trees, 15 % showed the greatest increased of LA (54.5 %) in respect to the control, and kept on decreasing as the addition levels in the silage increased, which could be due to the greater buffer capacity of the fodder trees, and could slightly decrease the efficiency of lactic bacteria (^{Lopez et al., 2008}). The addition of L. leucocephala and B. alicastrum in maize silages, produces a homolactic fermentation that ensures levels from organic acids, adequate for the conservation of the material.

The average content of N-NH₃/TN of all the treatments was 10.4 ± 2.9 %, all the treatments were found to be in accordance with ^{Ojeda et al. (2006)}, who report a range from 6.0 to 15.0 % of N-NH₃/TN, for guinea grass (P. maximum) added with white mulberry (M. alba). 42.9 % more ammoniacal nitrogen was detected in the silages of L. leucocephala (p<0.05), in comparison to silages of B. alicastrum; the silages with 30 % or more addition of fodder trees, contained 59.4 % more ammonia (p<0.05) than the rest of the treatments. In comparison to the control treatment, the silages of L. leucocephala and B. alicastrum produced 18.6 % and 43.5 % less ammonia, respectively. The mixed silages produced less ammonia, probably because pf the content of secondary compounds in the fodder species, as tannins, being able to form strong complexes with proteins and increase their degradation rate (^{Broderick, 1995}; ^{Mcsweeney et al., 1999}; ^{Silanikove et al., 2001}).

The mixed silages (gramineae + fodder trees) could represent an option to increase the input of nutrients in ruminant diets in the tropic. When adding L. leucocephala and B. alicastrum in maize silages (Z. mays), the content of protein is significantly increased and a predominantly homolactic fermentation is obtained, which procures adequate levels of pH, organic acids and low rates of degradation of nitrogenated compounds.