Introduction

Brazil produces a great diversity of roughage sources for the feeding of livestock. However, for dairy cow nutrition in confined systems, it is important that the selected roughage species are adapted to the edaphoclimate of the respective region in order to ensure sufficient quality and quantity. In this regard, storage and transport conditions are another important point to consider. According to Van Soest (1994), roughage is essential for feeding lactating cows as it is the main source of fibrous carbohydrates. These are mostly represented by neutral detergent fiber (NDF), which consists of cellulose, hemicelluloses, and other non-glycolic compounds. Reviewing literature from 1918 to 2017, Hall and Mertens (2017) emphasized that it is crucial to improve the understanding of the dynamics of carbohydrate fermentation in high-production dairy cows.

In cows, roughage is degraded by ruminal bacteria and as result of fermentation of fibrous carbohydrates (FC) and non-fibrous carbohydrates (NFC), microbial protein (MP) and volatile fatty acids (VFAs) are produced. These compounds are the main sources of protein and energy for the productive and reproductive performance of lactating dairy cows and of nutrients for the synthesis of milk components (LANZAS et al., 2007; WHITE et al., 2017a, 2017b).

Among the different roughage sources used in tropical regions, corn (Zea mays) whole plant silage is one of the main forages for dairy cows in confinement systems due to its ability to produce dry matter (DM), easy harvesting and fermentation in the silo, presence of digestible neutral detergent fiber (NDF), and high concentrations of liquid energy (ROBINSON et al., 2016; HENTZ et al., 2017; ZARDIN et al., 2017). In addition to corn silage, sugarcane (Saccharum officinarum) also stands out due to low production costs, high DM production per hectare, resistance to pests and diseases, and availability during the dry season (when there is typically a shortage of forage), but offers only a low protein content and shows low rumen degradation of the fibrous fraction (CASTRO et al., 2009; TEIXEIRA JÚNIOR et al., 2016; VALADARES FILHO et al., 2017). Another important source of roughage is the forage palms Opuntia spp. and Nopalea spp., which are adapted to the edaphoclimatic conditions of the semi-arid regions. This adaption and their low production costs make forage palms potential alternatives for cattle feeding during prolonged droughts (GALVÃO JÚNIOR et al., 2014; PEREIRA et al., 2017).

A substantial number of studies has been conducted to investigate the effect of different roughage sources on feeding of dairy cows. Employing scientific systematization techniques (meta-analyses), data of previous studies can be retrieved in order to critically re-examine the results and statistically analyze metadata to improve our understanding of the subject (SAUVANT et al., 2008; RABIEE et al., 2012). Compared to classical reviews, meta-analyses offer the considerable advantage of integrating data of previous research into collective analyses, and therefore provide quantitative conclusions (NORMAND, 1999; LOVATTO et al., 2007; MARÍN et al., 2015). In this study, we performed meta-analyses to evaluate the composition, voluntary intake, and digestibility of different diets, as well as milk yield and composition as a function of the source of roughage used in the diets of Holstein cows confined under experimental conditions in Brazil.

Material and Methods

A database was compiled by surveying scientific publications accessible on the sites of the Scientific Electronic Library Online (SciELO), the Brazilian Institute of Information in Science and Technology (IBICT), Google Scholar, and in the main periodicals of Agrarian Sciences. For database compilation, we considered studies published between January 2000 and December 2015, thus comprising the last sixteen years of research on feeding lactating cows under experimental conditions in confinement systems in Brazil, whose references can be consulted on thesis of (ALESSIO 2017).

The publications (papers, theses, and dissertations) were initially selected by reading the abstract. Subsequently, the selected studies were fully read, and the following information was extracted completely or in part for integration into the database: number of different roughages, cow breed, individual body weight (BW), days in milk (DIM), diet composition (roughage proportion; DCRP), DM, crude protein (CP), ether extract (EE), non-fibrous carbohydrates (CNF), neutral detergent fiber (NDF), acid detergent fiber (ADF), total digestible nutrients (TDN). Furthermore, we processed data on voluntary intake, apparent digestibility of nutrients, and milk yield and composition. The data were tabulated in a Microsoft Excel spreadsheet. The design of this study was based on the methods described by Normand (1999), Lovatto et al. (2007), and Sauvant et al. (2008).

The resulting database comprised 109 experiments with 424 treatments of 3,903 individual lactating Holstein cows. We only processed data of studies in which the only source of roughage was corn silage (DRCS), diets with a single source of roughage other than corn silage (DRNCS), or diets composed by the combination of two roughage (DTR). DRNCS, consisted of sugar cane in natura or as silage, elephant grass in natura or as silage, alfalfa pre-dried or as hay, grass hay, sorghum, and star grass silage. Diets containing a combination of two roughages were composed of the following combinations (followed by the average respective proportions): corn silage plus grass hay (38.5: 6.5), sunflower silage (66.0: 34.0; 34.0: 66.0), in natura sugar cane (29.8: 20.2), forage palm (32.6: 35.7), pre-dried ryegrass (30.3: 11.7), or cassava branch silage (30.0: 25.6); forage palm plus sorghum silage (38.3: 30.7), grass hay (49.8: 25.3), or sugarcane bagasse (47.2: 26.8); sugarcane silage plus grass hay (24.4: 14.5), and grass hay plus Napier grass silage (30.3: 27.5).

We analyzed BW, DIM, RP, DM, NFC, NDF, ADF, CP, EE, TDN, and net milk energy of lactation (NEL) of the diets. Voluntary intake is typically expressed as an absolute value (in kg/day), however, we calculated intake as a percentage of body weight (% of BW) and transformed it to grams per kilograms of metabolic weight (g/kg of BW^0.75), which is expressed as DFM, NFC, NDF, ADF, CP, EE, and TDN. For the apparent digestibility, we evaluated DM, NCF, NDF, CP, and EE. The performance of lactating cows was expressed as milk yield (kg/day), milk yield corrected for energy and protein (ECM), feed efficiency (FE), and milk fat, protein, and lactose content. The NEL (Mcal/kg) = 0.0245 x TDN (%) - 0.15, the ECM = (0.327 * milk yield) + (12.95 * fat content * milk yield/100) + (7.65 * milk protein/milk yield (100) and FE = milk yield (kg/day)/DM intake (kg/day) were calculated according to NRC 2001 (NRC, 2001).

The data were analyzed statistically using the MEANS and UNIVARIATE procedures and variance analyses, using the GLIMMIX procedure of SAS® statistical software (SAS INSTITUTE, 2002). To perform a comparison of means of the formed groups, the assumptions of normality and homogeneity of variances were tested (Shapiro-Wilk test and Levene’s test, respectively). This mostly revealed significant deviations from normal distribution and homogeneity of variances (P < 0.0001, each), as at least one of the assumptions was not met by any one variable. Therefore, we chose an analysis of variance using the GLIMMIX procedure, which considers the distribution of the response variable for the comparison of means, and uses the LSMEANS procedure. Means were compared using a Tukey-Kramer test with a probability threshold of 5%. For all analyzed variables, a normal distribution was applied, in which a connection between the mean of the observations and the systematic part is made by the identity function.

The variance analysis was based on the method of restricted maximum likelihood (REML) which considers independency of fixed effects and random effects of each observation. The probability density function of the observations is therefore reflected in the sum of the functions of each part. The random effects, related to the components of variance, eliminate the bias resulting from the loss of degrees of freedom in the estimation of the fixed effects of the model. The REML considers the loss of degrees of freedom as a function of the fixed effects, thereby providing non-biased estimators with minimum variance for balanced data. For unbalanced data (even when providing biased estimates), the REML is recommended, as is meta-analysis in which each variable can present a different number of observations. For the variance analyses of the present study, the type of roughage used in the diet (corn silage, only one roughage, or two roughages combined) was considered as a fixed effect, and the number of cows per treatment was used as a random effect. In general, the statistical model was based on the equation below:

Results

In the diet of lactating Holstein cows under experimental conditions in Brazil, DRCS produced higher RP compared to DRNCS, whereas DRNCS led to higher levels of DM, NCF, TDN, and NEL compared to DRCS and TR (Table 1). The DRCS and DTR led to higher intakes of DM, NDF, ADF, CP, and EE in kg/day, percent of LW and g/kg of BW^0.75, in relation to DRNCS (Table 2), which did not seem to correlate with NFC and TDN intake. The DRNCS and DTR showed higher NFC digestibility, but with reduced digestibility of NDF, compared to the DRCS. The DTR showed the highest digestibility for CP, but lower digestibility for EE, compared to the other diets, whereas the DRCS stood out for higher NDF digestibility (Table 3).

Table 1

Average ± standard error of the means of general information and diet composition as a function of roughage source used in the diet of Holstein cows under experimental conditions in Brazil.

1 DRCS = only source of roughage used in the diet was corn silage;2 DRNCS = diets with one source of roughage but not corn silage (sugar cane in natura or silage, elephant grass in natura or silage, alfalfa pre-dried or as a hay, hay of grass, sorghum and star grass silage);3 DTR = diets based on the combination of two roughages (corn silage plus hay of grass,  sunflower silage, in natura sugar, forage palm, pre-dried ryegrass, or cassava branch silage; forage palm plus sorghum silage, grass hay, sugarcane bagasse; sugarcane silage plus grass hay, and grass hay plus Napier grass). Different letters after averages in the same line indicate significant differences (Tukey-Kramer test, P < 0.05).

Table 2

Average ± standard error of the means of intake as a function of the source of roughage used in the diet of Holstein cows under experimental conditions in Brazil.

1 DRCS = only source of roughage used in the diet was corn silage;2 DRNCS = diets with one source of roughage but not corn silage (sugar cane in natura or silage, elephant grass in natura or silage, alfalfa pre-dried or as a hay, hay of grass, sorghum and star grass silage);3 DTR = diets based on the combination of two roughages (corn silage plus hay of grass,  sunflower silage, in natura sugar, forage palm, pre-dried ryegrass, or cassava branch silage; forage palm plus sorghum silage, grass hay, sugarcane bagasse; sugarcane silage plus grass hay, and grass hay plus Napier grass). Different letters after averages in the same line indicate significant differences (Tukey-Kramer test, P < 0.05).

Table 3

Average ± standard error of the means of nutrients digestibility as a function of the source of roughage used in the diet of Holstein cows under experimental conditions in Brazil.

1 DRCS = only source of roughage used in the diet was corn silage;2 DRNCS = diets with one source of roughage but not corn silage (sugar cane in natura or silage, elephant grass in natura or silage, alfalfa pre-dried or as a hay, hay of grass, sorghum and star grass silage);3 DTR = diets based on the combination of two roughages (corn silage plus hay of grass,  sunflower silage, in natura sugar, forage palm, pre-dried ryegrass, or cassava branch silage; forage palm plus sorghum silage, grass hay, sugarcane bagasse; sugarcane silage plus grass hay, and grass hay plus Napier grass). Different letters after averages in the same line indicate significant differences (Tukey-Kramer test, P < 0.05).

Cows that received DRCS and DTR were more efficient in terms of milk production and ECM, with higher FE and lactose content, but lower fat and protein content compared to the DRNCS. Lactating cows fed DRNCS showed higher fat and protein content with lower milk yield, ECM, FE, and lactose content compared to DRCS and DTR diets (Table 4).

Table 4

Average ± standard error of the means for milk yield and composition, feed efficiency as a function of the source of roughage used in the diet of Holstein cows under experimental conditions in Brazil.

1 DRCS = only source of roughage used in the diet was corn silage;2 DRNCS = diets with one source of roughage but not corn silage (sugar cane in natura or silage, elephant grass in natura or silage, alfalfa pre-dried or as a hay, hay of grass, sorghum and star grass silage)3 DTR = diets based on the combination of two roughages (corn silage plus hay of grass,  sunflower silage, in natura sugar, forage palm, pre-dried ryegrass, or cassava branch silage; forage palm plus sorghum silage, grass hay, sugarcane bagasse; sugarcane silage plus grass hay, and grass hay plus Napier grass).4 Milk yield corrected for energy and protein (ECM) ECM = (0.327 * milk yield) + (12.95 * fat content * milk yield/100) + (7.65 * milk protein *milk yield /100), and5 Feed efficiency = milk yield (kg/day)/DM intake (kg/day) (NRC, 2001). Different letters after averages in the same line indicate significant differences (Tukey-Kramer test, P < 0.05).

Discussion

Diet composition analyses showed that the DRCS allowed for an increase in RP as a function of energy density in corn silage (Table 1) when using a corn plant harvest stage in which the grain has a hard, farinaceous consistency with DM content varying from 30 to 35% (RESENDE et al., 2005). In this condition, the amount of digestible DM is maximal and the relative NDF proportion is reduced by the progressive increase in starch content due to grain filling (NUSSIO et al., 2001). In consequence, the energy density increases due to the presence of digestible NDF and the high concentration of energy in liquid components, which allows the reduction of concentrated foods with little compromise in the physiology and performance of lactating cows (SILVA et al., 2015). Moreover, harvesting the corn plant at a height of 0.47 meters above the ground yields the optimal relationship between quantity and quality of silage. In addition, this harvesting practice may achieve greater potential of economic returns and can thus help to finance management and fertilization strategies to avoid degradation of foraging systems, which is mainly caused by the negative potassium balance in the system (HÜLSE et al., 2017). According to Velho et al. (2014), corn silages containing high concentrations of energy usually hold large amounts of starch so that this feed can be considered both a roughage and a concentrate.

The DRNCS are mostly represented by sugar cane in natura or as silage. These have a higher DM content (which may be related to the higher amount of concentrate in the diet), followed by DRCS and then DTR, in which the use of forage palm combined with one more other roughages is predominant. The low DM content of DTR is due to in natura forage palm, which contains water to approximately 90%. Thus, diets containing a high proportion of forage palm have a high moisture content, which contributes to alleviate water shortages in prolonged droughts. Forage palm is fed to lactating cows in combination with other roughages in order to provide balanced diets. However, it is crucial to keep the DM content above a certain threshold in order to prevent conditions such as diarrhea and weight loss, which may occur when forage palm is fed as the only roughage source or ad libitum (GALVÃO JÚNIOR et al., 2014; PEREIRA et al., 2017).

When forage palm is used in a DTR, it is important to correct for CP contents, as this particular source of roughage has an average of 4.08% CP (VALADARES FILHO et al., 2017). Thus, the use of non-protein nitrogen sources such as urea can be helpful in order to synchronize the degradation rate of energy and protein in the rumen (NRC, 2001; FORBES, 2007), regarding the high NFC and low NDF content of the forage palm (VALADARES FILHO et al., 2017). In consequence, it is crucial for the formulation of effective diets to observe the NDF content. The respective recommendations are 19 to 21% of physically effective NDF (LEAN et al., 2014), or 75% of the total NDF (28 to 33%) in the diet should be derived from forage (NRC, 1989). Furthermore, it is recommended to increase two percent units of the total NDF content and to reduce two percentage units in the NFC content of the diet for each percentage unit of reduced NDF content from roughage (NRC, 2001). The physically effective NDF content for adequate ruminal functioning is quantified either based on particle size, considering only particles larger than 8 mm (ZEBELI et al., 2012), or based on the percentage of the total diet retained in a sieve of 4.0 mm mesh size, using a particle separator (KMICIKEWYCZ; HEINRICHS, 2015). The long fiber particles form a floating fiber blanket in the rumen, which is important for maintaining rumination and ruminal motility, saliva production, stability of ruminal pH and, in consequence, for milk fat content (MERTENS, 1997; ZEBELI et al., 2012).

The DRCS and DTR produced higher intakes of DM, NDF, ADF, CP, and EE, expressed in kg/day, % BW, and g/kg of BW^0.75, compared to DRNCS, however no difference was found in NDF content regarding function of roughage type, which may explain the intake differences, as NDF restricts the intake due to physical limitations of ruminal filling (MERTENS, 1994). However, the differences between DRCS and DRNCS and DTR can be explained by the lower digestibility of NDF for both, CP for DRNCS and EE for DTR. The reduction in nutrient digestibility, especially that of NDF, is associated with an increase of lignin (VAN SOEST, 1994; HALL, 2014), which is not degradable by ruminal bacteria and therefore increases the retention time of feed in the rumen and impairs nutrient utilization and intake of new meal (NRC, 2001; KRIZSAN; HUHTANEN, 2013; HUHTANEN et al., 2016). Thus, the rumen may contain residuals from previous feedings, which have lower degradation rates and are diluted by the newly ingested fast-fermenting feed, but the accumulated residuals continue to occupy space and thereby limit intake (VAN SOEST, 1994; TRAXLER et al., 1998).

In the diets fed to lactating dairy cows in subtropical and tropical systems, indigestible NDF must be included in the diet formulation, as NDF is one of the main factors determining food intake and also plays a significant role in rumen diet degradation and voluntary intake. This is particularly prominent in feed systems based on C₄ forages with pronounced lignification, which withstands microbial degradation (VAN SOEST, 1994; TRAXLER et al., 1998; HARPER; MCNEILL, 2015). Therefore, formulating diets based on NDF content without reference to indigestible NDF may affect voluntary intake, apparent digestibility, and metabolizable energy content of the diet (HARPER; MCNEILL, 2015). The formulation of DRNCS is predominantly characterized by including sugar cane, which has low levels of protein and shows low ruminal degradation of the fibrous fraction (CASTRO et al., 2009; TEIXEIRA JÚNIOR et al., 2016; VALADARES FILHO et al., 2017).

Lactating Holstein dairy cows fed with DTR showed similar milk yield and composition, ECM, and FE as cows fed with DRCS. In contrast, cows fed DRNCS, despite receiving a higher proportion of concentrate in the diet, lower roughage proportion, with higher NFC, TDN, and NEL content in the diet, exhibited inferior performance compared to the other diets, in regard of milk yield and composition, ECM, and FE, most likely because nutrient availability was compromised by the reduced digestibility of NDF and CP, which had a negative effect on nutrient intake.

The differences found in milk fat and protein contents confirm the well-known antagonism between milk yield and milk fat and protein content, which is referred to as a ‘dilution effect’ in milk production (MIGLIOR et al., 2007). Regarding milk production and lactose content, these differences are related to the digestibility of the nutrients that compromise the intake, with negative effects on the FE due to the high energy-dependency of lactose (RIGOUT et al., 2002; QIAO et al., 2005), which is the main determinant of milk production (ALLEN; PIANTONI, 2014).

Conclusion

Diets based on the combination of two roughages produce milk yield and composition comparable to diets based on corn silage, whereas diets with a single source of roughage other than corn silage (even with a higher proportion of concentrate) lead to inferior intake, milk yield, and feed efficiency due to lower nutrient utilization. Diets based on corn silage allow for a higher proportion of roughage in the diet due to the inherent energy density.

Acknowledgments

We would like to thank all the Brazilian researchers cited in the references of the database, the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Fundação de Apoio à Pesquisa Científica e Tecnológica do Estado de Santa Catarina (FAPESC), and the Programa de Bolsas de Monitoria de Pós-Graduação (PROMOP) for the scholarships granted to postgraduate students participating in the study.

References

ALESSIO, D. R. M. Produção e composição do leite em função da alimentação de vacas em condições experimentais no Brasil. 2017. Tese (Doutorado em Ciência Animal) - Universidade do Estado de Santa atarina, Lages.

ALLEN, M. S.; PIANTONI, P. Carbohydrate nutrition: managing energy intake and partitioning through lactation. Veterinary Clinics of North America - Food Animal Practice, Philadelphia, v. 30, n. 3, p. 577-597, 2014.

CASTRO, H. S. de; ANDRADE, L. A. de B.; BOTREL, E. P.; EVANGELISTA, A. R. Rendimentos agrícola e forrageiro de três cultivares de cana-de-açúcar (Saccharum spp.) em diferentes épocas de corte. Ciências e Agrotecnologia, Lavras, v. 33, n. 5, p. 1336-1341, 2009.

FORBES, J. M. Voluntary food intake and diet selection in farm animals. 2. ed. Wallingford, UK: CAB International, 2007. 461 p.

GALVÃO JÚNIOR, J. G. B.; SILVA, J. B. A.; MORAIS, J. H. G.; LIMA, R. N. Palma forrageira na alimentação de ruminantes: cultivo e utilização. Acta Veterinária Brasilica, Mossoró, v. 8, n. 2, p. 78-85, 2014.

HALL, M. B. Feed analyses and their interpretation. Veterinary Clinics of North America - Food Animal Practice, Philadelphia, v. 30, n. 3, p. 487-505, 2014.

HALL, M. B.; MERTENS, D. R. A 100-Year review: carbohydrates - characterization, digestion, and utilization. Journal of Dairy Science, Champaign, v. 100, n. 3, p. 10078-10093, 2017.

HARPER, K.; MCNEILL, D. The Role iNDF in the regulation of feed intake and the importance of its assessment in subtropical ruminant systems (the role of iNDF in the regulation of forage intake). Agriculture, Switzerland, v. 5, n. 3, p. 778-790, 2015.

HENTZ, F.; VELHO, J. P.; NÖRNBERG, J. L.; HAYGERT-VELHO, I. M. P.; HENZ, E. L.; HENN, J. D.; PERIPOLLI, V.; ZARDIN, P. B. Fractionation of carbohydrates and nitrogenous constituents of late-crop corn silages ensiled with different specific masses. Semina: Ciências Agrárias, Londrina, v. 38, n. 1, p. 491-502, 2017.

HUHTANEN, P.; DETMANN, E.; KRIZSAN, S. J. Prediction of rumen fiber pool in cattle from dietary, fecal, and animal variables. Journal of Dairy Science, Champaign, v. 99, n. 7, p. 5345-5357, 2016.

HÜLSE, J.; NEUMANN, M.; UENO, R. K; HEKER JUNIOR, J. C.; FIGUEIRA, D. N.; SANDINI, I. E.; MÜLLER, M. M. L.; HORST, R. H.; VIGNE, G. L. D. Nutrient balance in the soil and nutritive characteristics of maize silage cut at different heights. Semina: Ciências Agrárias, Londrina, v. 38, n. 6, p. 3779-3796, 2017.

KMICIKEWYCZ, A. D.; HEINRICHS, A. J. Effect of corn silage particle size and supplemental hay on rumen pH and feed preference by dairy cows fed high-starch diets. Journal of Dairy Science, Champaign, v. 98, n. 1, p. 373-385, 2015.

KRIZSAN, S. J.; HUHTANEN, P. Effect of diet composition and incubation time on feed indigestible neutral detergent fiber concentration in dairy cows. Journal of Dairy Science, Champaign, v. 96, n. 3, p. 1715-1726, 2013.

LANZAS, C.; SNIFFEN, C. J.; SEO, S.; TEDESCHI, L. O.; FOX, D. G. A revised CNCPS feed carbohydrate fractionation scheme for formulating rations for ruminants. Animal Feed Science and Technology, Amsterdam, v. 136, n. 3-4, p. 167-190, 2007.

LEAN, I. J.; GOLDER, H. M.; HALL, M. B. Feeding, evaluating and controlling rumen function. Veterinary Clinics of North America - Food Animal Practice, Philadelphia, v. 30, n. 3, p. 539-575, 2014.

LOVATTO, P. A.; LEHNEN, C. R.; ANDRETTA, I.; CARVALHO, A. D.; HAUSCHILD, L. Meta-análise em pesquisas científicas - Enfoque em metodologias. Revista Brasileira de Zootecnia, Viçosa, v. 36, suplemento especial, p. 285-294, 2007.

MARÍN, A. L. M.; SÁNCHEZ, N. N.; SIGLER, A. I. G.; BLANCO, F. P.; GARCÍA, V. D.; RUIPÉREZ, F. H. Metaanálisis del uso de semillas y aceites en la dieta de ovejas y cabras. Pesquisa Agropecuária Brasileira, Brasília, v. 50, n. 9, p. 821-828, 2015.

MERTENS, D. R. Creating a system for meeting the fiber requirements of dairy cows. Journal of dairy science, Champaign, v. 80, n. 7, p. 1463-1481, 1997.

MERTENS, D. R. Regulation of forage intake. In: FAHEY JUNIOR, G. C.; COLLINS, M.; MERTENS, D. R.; MOSER, L. E. (Ed.). Forage quality, evaluation and utilization. Madison: American Society of Agronomy, 1994. p. 450-493.

MIGLIOR, F.; SEWALEM, A.; JAMROZIK, J.; BOHMANOVA, J.; LEFEBVRE, D. M.; MOORE, R. K. Genetic analysis of milk urea nitrogen and lactose and their relationships with other production traits in Canadian Holstein cattle. Journal of Dairy Science, Champaign, v. 90, n. 5, p. 2468-2479, 2007.

NORMAND, S. T. Tutorial in biostatistics: meta-analysis: formulating, evaluating, combining and reporting. Statistics in Medicine, New York, v. 359, n. 18, p. 321-359, 1999.

NATIONAL RESEARCH COUNCIL - NRC. Nutrient requeriments of dairy cattle. 6. ed. Washington, DC: National Academy Press, 1989. 158 p.

______. ______. 7. ed. Washington D.C.: National Academy Press, 2001. 406 p.

NUSSIO, L. G.; CAMPOS, F. P. de; DIAS, F. N. Importância da qualidade da porção vegetativa no valor alimentício da silagem de milho.. In: JOBIM, C. C.; CECATO, U.; DAMASCENO, J. C.; SANTOS, G. T. SIMPÓSIO SOBRE PRODUÇÃO E UTILIZAÇÃO DE FORRAGENS CONSERVADAS, 2001, Maringá. Anais... Maringá: UEM/CCA/DZO, 2001. p. 127-145.

PEREIRA, P. de C.; SILVA, T. G. F. da; ZOLNIER, S.; SILVA, S. M. S.; SILVA, M. J. Water balance in soil cultivated with forage cactus clones under irrigation. Revista Caatinga, Mossoró, v. 30, n. 3, p. 776-785, 2017.

QIAO, F.; TROUT, D. R.; XIAO, C.; CANT, J. P. Kinetics of glucose transport and sequestration in lactating bovine mammary glands measured in vivo with a paired indicator/nutrient dilution technique. Journal of Applied Physiology, San Diego, v. 99, n. 3, p. 799-806, 2005.

RABIEE, A. R.; BREINHILD, K.; SCOTT, W.; GOLDER, H. M.; BLOCK, E.; LEAN, I. J. Effect of fat additions to diets of dairy cattle on milk production and components: a meta-analysis and meta-regression. Journal of Dairy Science, Champaign 95, n. 6, p. 3225-3247, 2012.

RESENDE, F. D. DE SIGNORETTI, R. D.; COAN, R. M.; SIQUEIRA, G. R. (Ed.). Terminação de bovinos de corte com ênfase na utilização de volumosos conservados.. In: REIS, R. A.; SIQUEIRA, G. R.; BERTIPAGLIA, L. M. A.; OLIVEIRA, A. P.; MELO, G. M. P.; BERNARDES, T. F. (Ed.). VOLUMOSOS NA PRODUÇÃO DE RUMINANTES, 2005. Jaboticabal. Anais... Jaboticabal: Fundação de Apoio à Pesquisa, Ensino e Extensão, FUNEP, 2005. p. 83-106.

RIGOUT, S.; LEMOSQUET, S.; VAN EYS, J. E.; BLUM, J. W.; RULQUIN, H. Duodenal glucose increases glucose fluxes and lactose synthesis in grass silage-fed dairy cows. Journal of Dairy Science, Champaign, v. 85, n. 3, p. 595-606, 2002.

ROBINSON, P. H. , SWANEPOEL, W.; HEGUY, J. M.; PRICE, T.; MEYER, D. M. Total ‘shrink’ losses in commercially sized corn silage piles: Quantifying total losses and where they occur. Science of the Total Environment, Amsterdam, v. 542, n. 4, p. 530-539, 2016.

STATISTICAL ANALYSIS SYSTEM - SAS. Statistical analysis system user’s guide: statistics. Version 8.2, Cary: SAS Institute, 2002.

SAUVANT, D.; SCHMIDELY, P.; DAUDIN, J. J.; STPIERRE, N. R. Meta-analysis of experimental data in animal nutrition. Animal, Cambridge, v. 2, n. 8, p. 1203-1214, 2008.

SILVA, M. S. J. da; JOBIM, C. C.; POPPI, E. C.; TRES, T.; OSMARI, M. P. Production technology and quality of corn silage for feeding dairy cattle in Southern Brazil. Revista Brasileira de Zootecnia, Viçosa, v. 44, n. 9, p. 303-313, 2015.

TEIXEIRA JÚNIOR, F. E. P.; RUAS, J. R. M.; LOPES, M. A.; COSTA, M. D. da; PIRES, D. A. de A.; ROCHA JÚNIOR, V. R. Effect of different alternative roughages on the profitability of milk production systems with F1 Holstein x Zebu cows. Revista Ceres, Viçosa, v. 63, n. 6, p. 807-815, 2016.

TRAXLER, M. J.; FOX, D. G.; VAN SOEST, P. J.; PELL, A. N.; LASCANO, C. E.; LANNA, D. D.; MOORES, J. E. ; LANA. R. P.; VÉLEZ, M.; FLORES, A. Predicting forage indigestible NDF from lignin concentration. Journal of Animal Science, Champaign, v. 76, n. 5, p. 1469-1480, 1998.

VALADARES FILHO, S. C.; MACHADO, P. A. S.; CHIZZOTTI, M. L. CQBAL 3.0. Tabelas Brasileiras de Composição de Alimentos para Bovinos. Disponível em: <www.ufv.br/cqbal>. Acesso em: 05 set 2017.

VAN SOEST, P. J. Nutritional ecology of the ruminant. Ithaca: Cornell University, 1994. 476 p.

VELHO, J. P.; MÜHLBACH, P. R. F.; GENRO, T. C. M.; BARCELLOS, J. O. J.; BRACCINI NETO, J.; SILVA, R. S. M. Modelos matemáticos para ajuste da produção de gases in vitro em diferentes tempos de incubação e cinética ruminal de silagens de milho. Semina: Ciências Agrárias, Londrina, v. 35, n. 4, p. 2531-2540, 2014.

WHITE, R. R.; ROMAN-GARCIA, Y.; FIRKINS, J. L.; KONONOFF, P.; VANDEHAAR, M. J.; ARMENTANO, L. E.; WEISS, W. P.; McGILL, T.; GARNETT, R.; HANIGAN, M. D. Evaluation of the national research council (2001) dairy model and derivation of new prediction equations. 1. Digestibility of fiber, fat, protein, and nonfiber carbohydrate. Journal of Dairy Science, Champaign, v. 100, n. 5, p. 3591-3610, 2017a.

WHITE, R. R.; ROMAN-GARCIA, Y.; FIRKINS, J. L.; KONONOFF, P.; VANDEHAAR, M. J.; TRAN, H.; McGILL, T.; GARNETT, R.; HANIGAN, M. D. Evaluation of the national research council (2001) dairy model and derivation of new prediction equations. 2. Rumen degradable and undegradable protein. Journal of Dairy Science, Champaign, v. 100, n. 5, p. 3611-3627, 2017b.

ZARDIN, P. B.; VELHO, J. P.; JOBIM, C. C.; ALESSIO, D. R. M.; HAYGERT-VELHO, I. M. P.; CONCEIÇÃO, G. M.; ALMEIDA, P. S. G. Chemical composition of corn silage produced by scientific studies in Brazil - A meta-analysis. Semina: Ciências Agrárias, Londrina, v. 38, n. 1, p. 503-511, 2017.

ZEBELI, Q.; ASCHENBACH, J. R.; TAFAJ, M.; BOGUHN, J.; AMETAJ, B. N.; DROCHNER, W. Invited review: Role of physically effective fiber and estimation of dietary fiber adequacy in high-producing dairy cattle. Journal of Dairy Science, Champaign, v. 95, n. 3, p. 1041-1056, 2012.

Notes

Author notes

* Author for correspondence