Previous Page Table of Contents Next Page

Poster 7.4: Tropical maize silage in central Brazil - Raúl R. Vera and Esteban A. Pizarro

Raúl R. Vera

Esteban A. Pizarro

Pontificia Universidad Católica de Chile
Santiago, Chile
E-mail: [email protected]

Universidad de la República
Montevideo, Uruguay


The long (5-7 month) dry season of the savannah region (cerrados) of central Brazil makes imperative the conservation of forage to maintain continuous production of milk.

Surveys conducted in the late 1970s and early 1980s indicated that in the State of Minas Gerais the conservation of maize, sorghum, elephant grass and mixtures of any two of these species were extremely common among dairy producers. Analysis of on-farm silage samples (Paiva et al., 1978) showed that the resulting silages were always of low nutritive value, with a mean in vitro organic matter digestibility (IVOMD) of 60% and 5.6% CP. A parallel survey (unpublished) also showed that the process of ensilage was slow and in small-scale dairy units it was frequently carried out manually.

It was therefore hypothesized that the low resulting nutritional value was due at least in part to the slow process of cutting, chopping and ensiling. A large number of experiments was carried out to determine maize growth curves; nutritional value throughout the vegetative period; various measurements of the efficiency (Pizarro and Vera, 1980) with which field operations can be carried out; material losses during cutting, chopping and conservation; and the resulting nutritional value of maize silages. Lastly, a computer model of the whole process and various alternatives was developed (Pizarro and Vera, 1979; Vera and Pizarro, 1981).

What follows is a brief summary of results, with emphasis on the inherent nutritional value of maize, and of the resulting silage, when the crop is grown under tropical conditions and on low-fertility oxisols.


Locally bred and widely available maize cultivars were used in all of the trials, including ‘Agroceres 259’, ‘Dentado Composto’, ‘BR103’ and ‘Maia 13’. The last two were varieties released by the EMBRAPA National Centre for Maize and Sorghum Research.

The growth curves of the two last-named cultivars were studied for three consecutive years, and regular samples were collected between days 23 and 170 post-planting. Total yields and components of the biomass were quantified.

The maize crops were fertilized according to the then current recommendations, including 100 kg/ha N, 40 P and 40 K. A typical soil analysis (0-20 cm) was: clay - 65%; sand - 13%; pH - 5.2; P - 2 ppm; organic matter - 2.6%.

Analysis of the growth curves showed that, for all practical purposes, between-year differences were accounted for by differences in accumulated temperature (ACCTEMP) and rainfall (ACCRAIN), and these two variables, together with days since planting (AGE), provided a good prediction of DM yields:

Yield = 8.22 ACCRAIN + 0.00080 ACCRAIN2 + 4.803 ACCTEMP - 52.402 AGE - 0.2212 AGE2 - 2659 (r = 0.96)

DM percentage of the whole plant (DM PC) did not vary significantly between years, and was largely accounted for by AGE:

DM PC = 7.66 (0.0120 AGE) (r = 0.95)

The most striking result in terms of nutritional value was the rapid decline in the CP content of the crop, regardless of year and variety. CP tended to stabilize at 4 to 5% after the 100th day of growth, as follows:

CP = 22.56 (-0.0285 AGE) + 6.09 (-0.003085 AGE)

DM digestibility (DMD) of the standing crop was evaluated in two sets of data. The first one determined the in vitro DMD of samples collected throughout the growth period, as explained above. The second set of data was derived from a continuous digestibility trial carried out with penned sheep between days 49 and 177 of the growth period; it should be noted that over the period 140-177 days, the crop was fully mature and field-dried.

Up until 140 days of age, digestibility decreased linearly:

DMD = 73.98 - 0.172 AGE (r = 0.84)

This implies that over the period of 100-120 days of age, which corresponds to approximately 30% DM (the stage generally recommended for ensilage), DMD would be roughly 50 to 55% and CP 5%.

At approximately this stage of maturity, the contribution of grain to total yield was unexpectedly low, despite being reasonably high in absolute terms, as shown in Table 1.

Table 1. Total DM and grain yields in two tropical maize cultivars

BR 105

Maia 13

DM yield (kg/ha)

11 626

18 078

Grain yield (kg/ha)

3 288

4 237

Grain yield (% of total yield)



As shown in Table 2, soluble carbohydrates, and starch in particular, were low in the fresh forage. The low starch content in the fresh forage, relative to the ensiled material, is almost certainly due to a laboratory artefact, since later data determined in a different laboratory found that at a comparable stage of growth, starch in the DM of fresh forage ranged between 18 and 19% (Neto et al., 1984). Nevertheless, soluble carbohydrates in the latter case were even lower than above.

Table 2. Chemical composition of the fresh material and the resulting silage of cv ‘BR 105’

Green forage


DM (%)



Ethanol-soluble carbohydrates (%DM)



Starch (%DM)



Cellulose (%DM)



Hemicellulose (%DM)



Lignin (%DM)



It is worth noting that Neto et al. (1984) analysed samples using “definitive” methods (Bailey, 1967, 1973) and were able to account for 85-90% of the DM, the remaining being ash and possibly minor fractions unaccounted for.

For the purpose of comparison, it should be noted that the expected composition of temperate maize is generally as follows: water soluble CHOs - 15%; starch - 25%; hemicellulose - 18%; cellulose - 23%; lignin - 5%; protein - 9%; DMD - 75%; and grain as % of total yield - 35-40%. A comparison with the data presented above shows that tropical maize in the experimental conditions reported here tends to be considerably higher in hemicellulose, somewhat higher in lignin and lower in protein and non-structural carbohydrates. It is hypothesized that this may be in part a plant adaptation to soil constraints, but mostly reflects the relatively low ratio of grain relative to the rest of the plant.

Numerous other results, particularly the partitioning of energy and nitrogen digestion in the gastro-intestinal tract of the animal, are available, but the above data should suffice to show the limitations of tropical maize silage, at least when grown on poor soils. In this environment, sorghum showed many of the same characteristics of maize (Pizarro et al., 1984), namely, high DM yields, moderate digestibility and marginal CP.

Not unexpectedly, the above tropical silages were unable to support weight gains in steers, unless supplemented with a protein supplement (Table 3).

Table 3. Liveweight gains of steers fed maize silage plus different levels of cottonseed cake

Cottonseed cake (kg/day/head)

LWG (kg/day/head)







Contrary to our initial hypothesis, it is clear that the low nutritional value of farm silages cannot be attributed to the speed with which field operations are carried out, since the crop is of low quality throughout a relatively long vegetative period, including stages earlier than those most appropriate for ensilage.


Bailey, R.W. 1967. Quantitative studies of ruminant digestion: loss of ingested plant carbohydrates from the reticulo-rumen. New Zealand J. Agric. Res., 10: 15-32.

Bailey, R.W. 1973. Structural carbohydrates. p. 157-211, in Vol. 1 of G.W. Butler & R.W. Bailey (eds) Chemistry and Biochemistry of Herbage. New York: Academic Press.

Neto, J.M., Vera, R.R., & Pizarro, E.A. 1984. Produção e avaliação qualitativa do milho dentado composto, cultivar BR 126. II. Valor nutritivo e digestibilidade in vitro. p. 380, in: 21st Reunião Anual da Sociedade Brasileira de Zootecnia. Belo Horizonte.

Paiva, J.A.J. de, Pizarro, E.A., Rodríguez, M., & de A.C. Viana, J. 1978. Qualidade da silagem ra regiao metalúrgica de Minas Gerais. Belo Horizonte, Brazil, Arquivos da Escola de Veterinaria da U.F.M.G., 30: 81-88.

Pizarro, E.A., & Vera, R.R. 1979. Efficiency of fodder conservation systems: maize silage. In: C. Thomas (ed) Forage Conservation in the 80s. British Grassland Society Occasional Symposium, No.11: 436-441.

Pizarro, E.A., & Vera, R.R. 1980. [Efficiency of maize harvesting for silage.] (in Portuguese). Belo Horizonte, Brazil, Arquivos da Escola de Veterinaria da U.F.M.G., 32: 127-130.

Vera, R.R., & Pizarro, E.A. 1981. [A mathematical model for the production of maize silage.] (in Portuguese). Belo Horizonte, Brasil, Arquivos da Escola de Veterinaria da U.F.M.G., 33: 553-567.

Pizarro, E.A., Vera, R.R., & Liseu, L.C. 1984. Curva de crecimiento y valor nutritivo de sorgos forrajeros en los trópicos. Prod. Anim. Trop., 9: 187-196.

Previous Page Top of Page Next Page