SciELO - Scientific Electronic Library Online

 
vol.48 issue2Design specifications of a heat applicator weed controller device for family farmsDetection of Campylobacter spp. in chilled and frozen broiler carcasses comparing immunoassay, PCR and real time PCR methods author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

  • text new page (beta)
  • English (pdf)
  • Article in xml format
  • How to cite this article
  • SciELO Analytics
  • Curriculum ScienTI
  • Automatic translation

Indicators

Related links

Share


Ciência Rural

On-line version ISSN 1678-4596

Cienc. Rural vol.48 no.2 Santa Maria  2018  Epub Feb 01, 2018

https://doi.org/10.1590/0103-8478cr20160855 

BIOLOGY

Nutritional diversity of Brachiaria ruziziensis clones

Divergência nutricional de clones de Brachiaria ruziziensis

Ellen de Almeida Moreira1 

Shirley Motta de Souza2 

Alexandre Lima Ferreira3 

Thierry Ribeiro Tomich2 

José Augusto Gomes Azevêdo1 

Fausto de Souza Sobrinho2 

Flávio Rodrigo Gandolfi Benites2 

Fernanda Samarini Machado2 

Mariana Magalhães Campos2 

Luiz Gustavo Ribeiro Pereira2  * 

1Universidade Estadual de Santa Cruz (UESC), Ilhéus, BA, Brasil.

2Embrapa Gado de Leite, 36038-330, Juiz de Fora, MG, Brasil.

3Universidade Federal de São João del Rei (UFSJ), São João del Rei, MG, Brasil.


ABSTRACT:

The aim of this study was to evaluate the nutritional diversity of Brachiaria ruziziensis clones through chemical composition and in vitro kinetics of ruminal fermentation. Twenty three clones of Brachiaria ruziziensis were used (15, 16, 46, 174, 411, 590, 651, 670, 768, 776, 844, 859, 950, 965, 970, 975, 1067, 1093, 1296, 1765, 1806, 1894 and 1972) and Brachiaria ruziziensis cv. ‘Kennedy’, Brachiaria brizantha cv. ‘Marandu’ and Brachiaria decumbens cv. ‘Basilisk’ as controls within 27 days of harvesting. The experimental design used randomized blocks with 26 treatments (genotypes) and three replications. Evaluation of the nutritional divergence was performed using principal components analysis, based on the following discriminatory variables: in vitro dry matter digestibility (IVDMD), neutral detergent fiber (NDF), lignin, crude protein (CP), degradation rate of non-fibrous carbohydrates (KdNFC) and degradation rate of fibrous carbohydrates (KdFC). The evaluation of the nutritional diversity of Brachiaria genotypes was based on the two main components (IVDMD and NDF), which explains 96.2% of the total variance Variables of lower contribution to the discrimination of the clones were as degradation rates of the fibrous and non-fibrous carbohydrates. In the agglomerative hierarchical grouping analysis, five distinct groups were identified, where V group, formed by clones 46, 768 and 1067 have higher values of IVDMD compared to the other clones.

Key words: digestibility; discriminating variables; main components; nutritional value

RESUMO:

O objetivo deste estudo foi avaliar a divergência nutricional de clones de Brachiaria ruziziensis através da composição química e cinética de fermentação ruminal in vitro. Os tratamentos consistiram de 23 clones de Brachiaria ruziziensis (15, 16, 46, 174, 411, 590, 651, 670, 768, 776, 844, 859, 950, 965, 970, 975, 1067, 1093, 1296, 1765, 1806, 1894 e 1972), e as testemunhas Brachiaria ruziziensis cv ‘Kennedy’, Brachiaria brizantha cv ‘Marandu’ e a Brachiaria decumbens cv ‘Basilisk’, colhidas com 27 dias de rebrota. O delineamento experimental utilizado foi o de blocos casualizados, com 26 tratamentos (genótipos) e três repetições. A avaliação da divergência nutricional foi realizada utilizando-se a análise de componentes principais e agrupamento aglomerativo hierárquico. Com base nas seguintes variáveis discriminatórias: digestibilidade in vitro da matéria seca; fibra em detergente neutro; lignina; proteína bruta; taxa de degradação de carboidratos não fibrosos e; taxa de degradação de carboidratos fibrosos. A avaliação da divergência nutricional dos clones de B. ruziziensis baseou-se nos dois primeiros componentes principais (DIVMS e FDN), explicando 96.2% da variância total. As variáveis de menor contribuição para a discriminação dos clones foram as taxas de degradação dos carboidratos fibrosos e não fibrosos. Na análise de agrupamento aglomerativo hierárquico foram identificados cinco grupos distintos, em que o grupo V, formado pelos clones 46, 768 e 1067 destacou-se em relação aos demais por apresentar valores superiores de digestibilidade in vitro da matéria seca.

Palavras-chave: componentes principais; digestibilidade; valor nutricional; variáveis discriminatórias

INTRODUCTION:

In Brazil, beef and dairy cattle production is based primarily on grass-feeding systems (around 90%), making pasture grasses the main source of animal feed. With growth in the livestock sector, the search for foods that combine high production and high-quality has been increasing. However, only a few varieties meet the requirements, demonstrating the importance of the introduction of genetically improved cultivars.

Although, the number of forage species available in Brazil is high, Brachiaria and Panicum occupy the largest area. Among the Brachiaria species cultivated in Brazil, Brachiaria ruziziensis is not widely used in the country despite showing promise for breeding programs with high nutritional quality, good adaptation in crop-livestock-forest integration system (CLFIS), suitable ground cover with direct planting, and as the only diploid sexual species, which allows variability between generations for selection of superior genotypes. However, this forage species also has some unfavorable characteristics, such as susceptibility to be attacked by spittlebug, with low productivity, and reduced adaptation to less fertile and acidic soils (DIAS et al., 2013).

Multivariate analysis has been used to evaluate nutritional diversity in forage species (AZEVÊDO et al., 2003; FREITAS et al., 2006), helping to identify genotypes with genetic differences that produce progeny with greater heterogeneity, thus increasing the likelihood of obtaining superior individuals in segregating generations (CRUZ et al., 2012; SHIMOYA et al., 2002).

Selecting for good performance and high nutritional value improves efficiency of a breeding program (CRUZ et al., 2012). In this context, the objective was to evaluate the nutritional diversity of 23 Brachiaria ruziziensis clones in the breeding program at EMBRAPA.

MATERIALS AND METHODS:

The experiment was performed at the Experimental Complex Multiuser of Bioefficacy and Sustainability of Livestock and at the experimental farm of José Henrique Bruschi of EMBRAPA Dairy Cattle in Coronel Pacheco-MG, Brazil (23°35’16” S, 43°15’56” W, altitude of 426m). The climate corresponds to the Cwa type (mesothermal) in the Koppen classification and the soil of the experimental area is classified as Red-Yellow Alic Argisol (SANTOS et al., 2006).

The experimental design used randomized blocks with 26 treatments (genotypes) and three replications. Treatments consisted of 23 clones of Brachiaria ruziziensis, from the forage breeding program of EMBRAPA, represented by IDs: 15, 16, 46, 174, 411, 590, 651, 670, 768, 776, 844, 859, 950, 965, 970, 975, 1067, 1093, 1296, 1765, 1806, 1894 and 1972. Brachiaria ruziziensis cv. ‘Kennedy’, Brachiaria brizantha cv. ‘Marandu’ and Brachiaria decumbens cv. ‘Basilisk’ were used as a control.

Plants in each plot were cut within 27 days of growth, at an average height of 10cm with the aid of motorized costal mower. After collection, samples were weighed fresh and drying ovens with forced air circulation at 65°C for 72 hours. They were ground in a Wiley mill with 1 mm sieve and stored in polyethylene bottles for later composition analyses.

The contents of dry matter (DM), crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), lignin, and in vitro dry matter digestibility (IVDMD), were determined by near infra-red spectroscopy (NIRS) in the ruminant nutrition laboratory at the Animal Science section of EMBRAPA Dairy Cattle, Juiz de Fora, MG, Brazil.

The in vitro rumen fermentation kinetics was performed by the semiautomatic gas production technique following the procedures described in MAURICIO et al. (1999). For this evaluation, 50mL glass flasks were utilized and 0.32g of the substrate to be tested were incubated in filter bags F57 (Ankon®). Twenty-eight mL of buffered culture medium and rumen fluid, prepared the previous day (MENKE & SEINGASS, 1987) and kept pressurized under CO2, were added. Flasks were sealed with a silicone stopper to avoid contamination and fermentation and refrigerated at 4ºC. Five hours before inoculation with rumen fluid, the bottles were placed in a room at 39°C.

The rumen inoculate was collected from three fistulated Holstein x Gyr dry cows with 500±15kg average body weight. Diet comprised of pasture (Brachiaria decumbens) supplemented with 15kg/day of corn silage and 2.0kg of concentrate a day. Rumen fluid was collected in the morning before feeding through the ruminal cannula and then filtered through double layers of cheesecloth and maintained at 39°C. Inoculum of three cows was pooled.

Three milliliters of rumen fluid was added to the flasks containing samples and buffered culture medium. Finally, flasks were sealed with a silicone stopper and aluminum washers to avoid gases escaping. Triplicates of each sample were incubated and kept heated at 39ºC room. Flasks containing only inoculum and culture medium were used as a blank.

Pressure readings were taken using a pressure transducer (DPI 705 - GE) at 2, 4, 6, 8, 10, 12, 14, 17, 20, 24, 28, 34, 48, 72 and 96 hours after inoculation. The PSI values were converted to volume according to the equation: Volume = -0.0125x2+3.6015x - 0.1118; R²=0.9874, established from the laboratory conditions. Production volumes were adjusted to g of substrate (based on DM) incubated and the values obtained were corrected for blanks (flasks without substrate at each incubation time).

A mathematical description of rumen fermentation kinetics was estimated using in vitro gas production. The bi-compartmental model was used and fitted to the curve of cumulative gas production (SCHOFIELD et al., 1994) as described below:

V (t) = VFNFC/(1+exp (2-4*kdNFC*(T-L)))+VFFC/ (1+exp (2-4*kdFC*(T-L)))

Where: V (t) = total gas accumulated at time t; VFNFC is equivalent to the maximum volume of gases from the NFC fraction (mL); VFFC is the maximum volume of the gases from the FC fraction (mL); kdNFC is the degradation rate (% h) of NFC; kdFC is the degradation rate (% h) of FC; and T and L are the incubation (hours) and lag (hours) times, respectively.

The data of bromatological composition were submitted to the univariate analysis through the Minitab 16 program and the averages compared by the Tukey test at 5% of probability. Principal components analyses and agglomerative hierarchical clustering (complete linkage) were conducted in Minitab 16 to evaluate the nutritional divergence between genotypes. Euclidean distance using standardized mean was used as a basic measure of similarity.

RESULTS AND DISCUSSION:

The dry matter (DM), in vitro dry matter digestibility (IVDMD), and crude protein (CP) variables had a significant effect among the evaluated clones (P<0.05; Table 1). The DM content ranged from 202.6 to 147.0g/kg DM, with the highest DM content observed for clone 859, with a mean of 202.6g/kg DM, whereas the clone presented lower IVDMD, with a mean of 577.5g/kg DM (P<0.05). The highest values of IVDMD were observed for clones 16, 46, 768, 970 and 1067, with a mean of 679.7 to 692.1g/kg DM (P<0.05). Clones showed high protein content (range 164.9 to 130.4g/kg DM) and clone 15 had the highest value of protein.

Table 1 Chemical compositions (g/kg DM) of the 23 clones and Brachiaria ruziziensis cv. ‘Kennedy’, Brachiaria brizantha cv. ‘Marandu’ and Brachiaria decumbens cv. ‘Basilisk’. 

Clones DM (g/kg) SD IVDMD (g/kg) SD NDF (g/kg) SD ADF (g/kg) SD Lignin (g/kg) SD CP (g/kg) SD
15 162.4abc 7.1 661.5ab 7.0 602.8a 13.3 306.9a 10.3 44.7a 4.1 164.9a 19.2
16 160.8abc 15.0 692.1a 29.4 580.4a 6.9 288.5a 4.1 43.0a 1.0 173.5ab 18.0
46 186.8abc 16.7 685.0a 36.1 611.4a 18.5 319.8a 26.4 45.4a 6.3 158.6abc 20.8
174 147.9c 34.3 637.7abc 16.8 630.0a 11.1 337.8a 3.3 46.4a 1.4 151.4abcd 5.8
411 159.8abc 12.6 655.7abc 34.4 639.5a 27.9 339.3a 25.0 40.6a 2.3 136.8bcd 21.2
590 159.0abc 21.1 629.6abc 27.0 643.1a 30.9 358.9a 29.4 44.2a 1.1 130.4d 11.8
651 167.1abc 7.2 652.8abc 14.7 617.1a 28.8 332.2a 38.7 46.7a 2.1 140.8abcd 18.5
670 165.6abc 11,6 626.0abc 31.0 609.0a 33.3 342.4a 23.7 53.5a 7.0 156.6abcd 23.8
768 151.7abc 7.5 690.2a 40.2 605.3a 14.2 311.7a 12.5 42.9a 1.9 148.7abcd 5.9
776 151.4abc 3.3 647.9abc 28.8 617.9a 13.6 346.3a 16.8 45.2a 3.1 144.0abcd 26.1
844 173.5abc 13.0 593.3bc 17.7 616.8a 15.9 324.7a 10.9 47.6a 5.0 149.4abcd 17.1
859 202.6a 31.4 577.5c 15.8 608.2a 46.9 329.8a 19.8 47.3a 5.0 153.7abcd 22.7
950 206.3ab 8.9 638.7abc 7.0 590.2a 12.6 300.0a 1.8 49.0a 13.8 168.3abcd 19.3
965 172.7abc 3.7 667.1ab 15.3 616.0a 8.0 319.3a 14.3 39.0a 2.5 148.1abcd 14.6
970 180.7abc 1.9 679.7a 31.9 602.1a 18.7 313.4a 13.7 46.1a 0.5 155.4abcd 12.4
975 148.2bc 7.7 640.2abc 18.5 635.3a 28.6 336.7a 28.3 42.1a 7.8 138.8abcd 7.8
1067 184.1abc 27.3 683.7a 68.3 623.7a 51.1 331.2a 58.8 43.6a 5.1 145.8abcd 29.7
1093 166.2abc 8.8 638.2abc 15.9 630.9a 27.6 340.9a 24.4 52.0a 3.3 160.4ab 23.4
1296 148.4bc 21.9 640.8abc 29.0 637.1a 33.2 354.6a 27.9 45.2a 3.5 143.3abcd 27.4
1765 163.3abc 6.3 647.5abc 43.7 631.9a 39.0 334.1a 31.4 45.1a 1.8 149.5abcd 22.5
1806 147.0c 10.5 637.8abc 16.4 618.5a 24.0 320.4a 24.5 47.7a 4.6 154.5abcd 14.3
1894 172.2abc 24.7 651.6abc 36.6 624.7a 21.5 331.5a 28.1 42.2a 7.6 141.4abcd 25.3
1972 160.2abc 21.7 644.9abc 12.1 628.9a 23.5 347.8a 25.1 48.9a 0.1 132.4cd 17.4
Ruziziensis 151.1abc 6.9 633.9abc 24.8 618.9a 20.4 343.6a 23.7 44.1a 5.7 141.8abcd 26.0
Brizantha 173.7abc 14.5 629.2abc 19.0 631.6a 13.1 342.8a 21.1 48.2a 2.9 150.4abcd 12.8
Decumbens 187.9abc 17.5 649.8abc 14.0 628.7a 28.9 334.8a 21.4 42.7a 3.7 137.7bcd 10.1
Average 166.9 646.9 620.1 331.3 45.5 148.5
SD 20.7 35.1 25.8 25.8 5.0 18.8
P-value 0.001 0.000 0.110 0.016 0.123 0.000

SD = Standard deviation of the mean; Means followed by the same letter within Column, do not differ by Tukey test (P<0.05).

No difference (P>0.05) was observed in neutral detergent fiber (NDF), acid detergent fiber (ADF) and lignin (Lig). The mean value of Lig was 45.5g/kg DM. For the NDF and ADF, the variations were 643.1 to 580.4 and 358.9 to 288.5g/kg of DM, respectively. According to VAN SOEST (1994), the NDF content influences the consumption of bulky foods and all clones presented values higher than 550g/kg, considered by this author as a limit to influence the consumption of forage.

LOPES et al. (2010) evaluated the nutritional quality of four Brachiaria species (B. brizantha, B. humidicula, B. decumbens and B. ruziziensis) at 56 days of growth and reported mean values of DM, NDF and ADF higher than those observed in the present study, 20.55, 68.44 and 36.53%, respectively, while mean values of CP, IVDMD and Lig were lower (6.9, 61.6 and 3.3%, respectively). Higher values of NDF and ADF and lower values of IVDMD and CP were also observed by SOUZA SOBRINHO et al. (2011) (NDF, ADF, IVDMD and CP of 78.0, 43.1, 55.2 and 6.3%, respectively) were evaluated for the forage quality of different species of Brachiaria, cut at 57 days of regrowth. Difference observed between the studies can be attributed to the stage of forage maturity, since digestibility and CP tend to decrease with the advancement of plant maturity and increase of the NDF and ADF fraction.

Table 2 shows the parameters of ruminal fermentation kinetics of B. ruziziensis clones and controls. Cumulative gas production rate for the fermentation of fibrous and non-fibrous carbohydrates (VFFC and VFNFC) ranged from 60.9 to 170.2mL/g and from 42.8 to 125.4mL/g, respectively. Volume of gas produced depends on the composition of the food and, the larger the amount of fiber, the greater the gas production (NOGUEIRA et al., 2006). This allowed that genotypes that presented superior VFFC are more digestible compared to the fibrous fraction. Structural carbohydrates have slower degradability and for this reason, the fibrous carbohydrate degradation rate (KdFC) is lower than the non-fibrous carbohydrate degradation rate (KdNFC), 0.019 and 0.052DM/h, respectively.

Table 2 Average of the adjusted parameters in relation to the in vitro fermentation kinetics of the fibrous carbohydrates (FC) and non-fibrous carbohydrates (NFC) in relation to the Brachiaria ruziziensis progenies and the Brachiaria ruziziensis cv. ‘Basilisk’. 

--------------------------------------------------------------Variables---------------------------------------------------------------
Clones VFNFC (ml/g-1) SE KdNFC (h-1) SE L (h:min) SE VFFC (ml/g-1) SE KdFC (h-1) SE
15 57.8 18.6 0.068 0.02 5:24 0.6 134.8 17.8 0.025 0.00
16 62.3 23.2 0.059 0.01 5:30 0.6 133.0 22.2 0.023 0.00
46 92.8 28.3 0.051 0.01 4:47 0.7 123.7 26.9 0.020 0.00
174 118.7 19.9 0.042 0.00 8:12 0.5 79.3 17.6 0.015 0.00
411 85.2 22.8 0.053 0.01 6:07 0.1 126.8 21.6 0.020 0.00
590 92.7 24.4 0.048 0.01 6:28 0.5 115.4 23.2 0.020 0.00
651 49.7 18.8 0.063 0.02 5:33 0.7 142.0 17.8 0.022 0.00
670 73.6 22.5 0.054 0.01 5:15 0.7 114.9 21.4 0.021 0.00
768 86.9 20.6 0.053 0.01 6:12 0.5 98.5 19.7 0.021 0.00
776 11.9 19.4 0.040 0.00 9:04 0.4 66.1 17.1 0.015 0.00
844 85.6 26.9 0.037 0.01 8:33 0.6 69.3 24.3 0.015 0.00
859 93.7 17.3 0.049 0.01 6:10 0.4 80.3 16.3 0.019 0.00
950 61.7 15.7 0.064 0.01 3:55 0.6 125.9 15.0 0.023 0.00
965 49.4 13.1 0.088 0.02 4:58 0.7 170.2 12.6 0.025 0.00
970 42.8 11.5 0.091 0.02 5:35 0.7 165.9 11.1 0.025 0.00
975 101.5 29.4 0.041 0.01 6:06 0.5 87.8 27.7 0.018 0.00
1067 88.4 36.9 0.043 0.01 6:44 0.8 75.5 34.9 0.018 0.01
1093 79.0 21.4 0.042 0.01 6:18 0.6 74.7 19.7 0.017 0.00
1296 122.9 28.9 0.036 0.00 6:27 0.6 60.9 25.4 0.014 0.01
1765 111.2 37.8 0.036 0.01 6:15 0.7 75.7 34.6 0.016 0.01
1806 60.1 16.4 0.074 0.02 5:08 0.6 135.2 15.7 0.025 0.00
1894 83.5 21.0 0.050 0.01 4:25 0.6 119.4 19.8 0.019 0.00
1972 125.4 21.5 0.038 0.00 5:47 0.5 60.7 19.1 0.015 0.00
B. ruziziensis 91.3 32.7 0.044 0.01 5:56 0.7 108.2 31.0 0.019 0.00
B. brizantha 91.9 22.2 0.046 0.01 5:39 0.6 95.4 20.6 0.018 0.00
B. decumbens 87.4 21.8 0.048 0.01 6:55 0.6 105.4 20.3 0.018 0.00
Average 85.0 0.052 6:03 105.6 0.019

VFNFC = equivalent to the maximum volume of gases from the NFC fraction (ml/g-1); kdNFC = the degradation rate (h-1) of NFC; VFFC = maximum volume of the gases from the FC fraction (ml/g-1); kdFC = degradation rate (h-1) of FC; L = lag time (hours:minutes); V(t) = total gas accumulated at time t and SE = Standard error.

The estimated colonization time (L) presented an average value of 6h. This parameter indicated the time involved between the beginning of the incubation and the start of microbial action on the sample. Thus, the greater amount of readily fermentable substances and the physical and chemical characteristics of the sample’s cell wall, which facilitated microbial colonization, represented lower time of colonization (MAGALHÃES et al., 2006).

Assessing the kinetic parameters of ruminal degradation of the fibrous and non-fibrous carbohydrate fractions of Brachiaria brizantha cv. ‘Marandu’ at three ages of cuts (28, 35 and 54 days) by the in vitro technique of gas production, SÁ et al. (2011), reported mean values for VFNFC and KdNFC at 28 days of age of 93.51mL/g of DM and 0.05h-1, respectively, values similar to those reported in the present study. However, the mean values observed for VFFC and KdFC (83.25mL/g DM and 0.01DM/h-1) were lower than those observed in the present study.

The evaluation of the nutritional divergence of Brachiaria genotypes (Table 3), was based on a principal component analysis (PC) where the cumulative variance of the first two principal components (PC1: 70.16% and PC2: 26.08%) explained 96.24% of variance between genotypes. Initially, we used all the variables of chemical and kinetic composition of fermentation, taken from minor models for discrimination of genotypes.

Table 3 Estimates of the eigenvalues of the cumulative variance and weighting of variables of the characters in the principal components obtained based on six variables in 26 genotypes of Brachiaria. 

-------------------------Weighting of variable----------------------------
Principal Component Variance (eigen value) Percentage of Total Variance (%) Variance Cumulative (%) IVDMD NDF Lignin CP KdNFC KdFC
1 7.5142 70.16 70.16 0.9620 -0.2400 -0.0476 0.1213 0.0023 0.0006
2 2.7933 26.08 96.24 0.2627 0.7973 -0.0802 -0.5375 -0.0031 -0.0009
3 0.3389 3.16 99.40 0.0446 0.5514 0.1883 0.8115 -0.0037 -0.0008
4 0.0641 0.60 100.00 0.0598 -0.0525 0.9776 -0.1945 -0.0068 -0.0019
5 0.0002 0.00 100.00 -0.0009 0.0048 0.0074 -0.0003 0.9822 0.1874
6 0.0000 0.00 100.00 0.0000 0.0003 0.0006 -0.0002 -0.1874 0.9823

IVDMD = in vitro dry matter digestibility ; NDF = neutral detergent fiber; CP = crude protein; kdNFC = degradation rate (h-1) of NFC and kdFC = degradation rate (h-1) of FC.

According to CRUZ et al. (2012), the relative importance of the main components decreases from the first to the last; the last component is responsible for explaining a tiny fraction of the total variance available. Thus, it was reported that the variables of lesser interest were KdNFC and KdFC once had a higher weighting in the smallest eigenvalue component (Table 3).

From the eigenvectors associated with the main components, we obtained the scores of the 26 Brachiaria genotypes. Graphical dispersion of the scores of the two main components can be reported in figure 1, where the distance of these points is proportional to the degree of dissimilarity between populations. Genotype grouping was observed in four sets. The first with clones 46, 768, 965, 970 and 106, the second one with clones 15, 16 and 950, the third one with clones 670, 844 and 859 and finally the fourth set with clones 174, 411, 590, 651, 776, 975, 1093, 1296, 1765, 1806, 1894, 1972 and the Brizantha, Decumbens and Ruziziensis controls. Clones 15, 16, 670, 844, 859 and 950 showed the highest dispersion of scores in the first two main components and were considered the most dissimilar.

Figure 1 Dispersion diagram of Brachiaria clones obtained by principal component analysis. 

For the hierarchical clustering analysis by full connection method, based on Euclidean distance average, we used six variables selected from the PC analysis (IVDMD, NDF, lignin, CP, KdNFC and KdFC) and obtained five distinct groups (Table 4). Group V, formed by clones 46, 768 and 1067, showed better results in relation to the others, with a higher average of IVDMD (686.3g/kg), which was the main discriminating factor. Group IV constituted by clones 15, 16 and 950 is distinguished by the low NDF, high CP contents and high degradation rate of the fibrous fraction.

Table 4 Groups of genotypes of Brachiaria and average variables in each group formed by the hierarchical agglomerative clustering method of Complete Linkage, based on standardized average Euclidean distance. 

Items --------------------------------------------------------Group----------------------------------------------------------
I II III IV V
Clones
174 670 651 15 46
411 844 965 16 768
590 859 970 950 1067
776 1093 1806
975 B. brizantha
B. ruziziensis
1296
1765
1894
1972
B. decumbens
IVDMD (%DM) 643.6 612.8 659.4 664.1 686.3
NDF (%DM) 630.5 619.3 613.4 591.1 613.5
Lignin (% DM) 44.2 49.7 44.9 45.5 44.0
CP (%DM) 140.7 154.1 149.7 168.9 151.0
KdNFC (h-1) 0.043 0.045 0.079 0.063 0.049
KdFC (h-1) 0.017 0.018 0.024 0.024 0.020

IVDMD = in vitro dry matter digestibility ; NDF = neutral detergent fiber; CP = crude protein; KdNFC = degradation rate (h-1) of NFC and KdFC = degradation rate (h-1) of FC.

Group V clones have IVDMD values higher than Group I and II that include traditional cultivars already consolidated in the domestic market, indicating the potential of the nutritional value of the members of this group for the breeding program. Group I introduced many clones and also included B. decumbens and B. ruziziensis (Kennedy). Group II, consisting of clones 670, 844, 859, 1093 and B. brizantha, showed the lowest IVDMD (612.8g/kg) and higher lignin content (49.7g/kg). Clones of this group have similar features to B. brizantha, a species of great importance in Brazilian livestock rearing.

CONCLUSION:

B. ruziziensis clones showed nutritional divergence and clones 46, 768 and 1067 were distinguished clones of high nutritional value. Clones 15, 16 and 950 are distinguished by the lower values of NDF and high protein levels. The divergent nutritional characteristics can guide new crosses in the breeding program complementing the agronomic parameters for the generation of superior genotypes.

BIOETHICS AND BIOSSECURITY COMMITTEE APPROVAL

All the management procedures of the animals were conducted according to ethical principles of animal experimentation, established by the Brazilian College of Animal Experimentation and the current legislation was approved by the Ethics Committee for Animal Use of EMBRAPA Dairy Cattle (No. 03/2014).

ACKNOWLEDGEMENTS

The authors acknowledge the Embrapa Gado de Leite, Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq - Edital Repensa - Projeto PECUS-RumenGases) for the financial support and to Fundação de Amparo à Pesquisa do Estado da Bahia (FAPESB) for the granting scholarship.

REFERENCES:

AZEVÊDO, J. A. G. et al. Evaluation of the Nutritional Divergence of the Sugarcane (Saccharum spp.) Varieties. Revista Brasileira de Zootecnia, v. 32, n. 6, p. 1431-1442, 2003. Available from: http://dx.doi.org/10.1590/S1516-35982003000600018 >. Accessed: June 12, 2016. doi: 10.1590/S1516-35982003000600018. [ Links ]

CRUZ, C. D. et al. (2012) Biometric Models Applied to Genetic Improvement. 4th Edition, UFV, Vicosa. [ Links ]

DIAS, K. O. DAS G. et al. Plot size and border effect on breeding of Urochloa ruziziensis. Pesquisa Agropecuária Brasileira, v. 48, n. 11, p. 1426-1431, 2013. Available from: http://dx.doi.org/10.1590/S0100-204X2013001100002 >. Accessed: June 21, 2016. doi: 10.1590/S0100-204X2013001100002. [ Links ]

FREITAS, W. P. et al. Evaluation of the nutritional divergence of sugarcane (Saccharum spp.) genotypes. Revista Brasileira de Zootecnia , v. 35, n. 1, p. 229-236, 2006. Available from: http://dx.doi.org/10.1590/S1516-35982006000100029 >. Accessed: June 12, 2016. doi: 10.1590/S1516-35982006000100029. [ Links ]

LOPES, F. C. et al. Chemical composition and in situ ruminal degradability of four Brachiaria species. Arquivo Brasileiro de Medicina Veterinária e Zootecnia, v. 62, n. 4, p. 883-888, 2010. Available from: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0102-09352010000400018 >. Accessed: June 12, 2016. doi: 10.1590/ S0102-09352010000400018. [ Links ]

MAGALHÃES, R. T. et al. Evaluation of four sorghum genotypes using the semi automated in vitro gas production. Revista Brasileira Milho e Sorgo, v. 5, n. 1, p.101-111, 2006. Available from: http://dx.doi.org/10.18512/1980-6477/rbms.v5n1.p101-111 >. Accessed: June 21, 2016. doi: 10.18512/1980-6477/rbms.v5n1.p101-111. [ Links ]

MAURÍCIO, R. M. et al. A semi-automated in vitro gas production technique for ruminants feedstuff evaluation. Animal Feed Science and Technology, v. 79, n. 4, p. 321-330, 1999. Available from: https://doi.org/10.1016/S0377-8401(99)00033-4 >. Accessed: Jan. 21, 2016. doi: 10.1016/S0377-8401(99)00033-4. [ Links ]

MENKE, H. H.; STEINGASS, H. Estimation of the energetic feed value obtained from chemical analysis and in vitro gas production using rumen fluid. Animal Research and Development, 28, 7-55, 1988. [ Links ]

NOGUEIRA, U.T. et al. Comparison among substrates with different soluble carbohydrates concentration using the in vitro semi-automatic gas production technique. Arquivo Brasileiro de Medicina Veterinária e Zootecnia , v. 58, n. 4, p. 633-641, 2006. Available from: http://dx.doi.org/10.1590/S0102-09352006000400027 >. Accessed: July 15, 2016. doi: 10.1590/S0102-09352006000400027. [ Links ]

SÁ, J. F. DE et al. In vitro ruminal fermentation kinetics of ‘Marandu’ grass at different harvest ages. Acta Scientiarum Animal Sciences, v. 33, n. 3, 2011. Available from: http://dx.doi.org/10.4025/actascianimsci.v33i3.9462 >. Accessed: July 19, 2016. doi: 10.4025/actascianimsci.v33i3.9462. [ Links ]

SANTOS, H. G. dos et al. (2006) Brazilian system of soil classification. 2th Edition. Rio de Janeiro: Embrapa Solos. [ Links ]

SCHOFIELD, P. et al. Kinetics of fiber digestion from in vitro gas production. Journal Animal Science. v. 72, n. 11, p. 2980-2991, 1994. Available from: https://www.animalsciencepublications.org/publications/jas/articles/72/11/2980 >. Accessed: July 25, 2016. doi: 10.2527 / 1994.72112980x. [ Links ]

SHIMOYA, A. et al. Genetic divergence among accessions of a germplasm bank of elephant grass. Pesquisa Agropecuária Brasileira , Brasília, v. 37, n. 7, p. 971-980, 2002. Available from: http://dx.doi.org/10.1590/S0100-204X2002000700011 >. Accessed: July 25, 2016. doi: 10.1590/S0100-204X2002000700011. [ Links ]

SOUZA SOBRINHO, F. DE et al. Productivity and quality of Brachiaria fodder in the Northern Region of Fluminense. Applied Research & Agrotechnology, v. 2, n. 3, p. 7-20, 2009. Available from: http://200.201.10.18/index.php/repaa/article/view/1502 >. Accessed: July 25, 2016. [ Links ]

VAN SOEST, P.J. Nutritional ecology of the ruminant. Ithaca, NY (USA): Cornell University Press, 1994. [ Links ]

0CR-2016-0855.R1

Received: September 17, 2016; Accepted: November 17, 2017; Revised: January 12, 2017

E-mail: luiz.gustavo@embrapa.br. *Corresponding author

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License