Introduction

The dry season of the year is considered the most critical phase of the grazing cattle production system. In this period, cattle consume forage of low nutritive value, characterized by high levels of indigestible fiber, and crude protein lower than the critical value of 7% DM [1]. Under such conditions, animals are subjected to multiple nutritional deficiencies, and protein (or nitrogen compounds) assumes a priority role. The available food or endogenous nitrogen recycling does not meet the minimum level of 10 mg/dL of ammoniacal nitrogen required to maintain the growth of cellulolytic bacteria [2], thereby promoting limitation in their activity and decrease in the cell wall digestion, leading to reduction of dry matter intake, digestibility of nutrients, and animal performance.

Therefore, strategies such as the provision of protein supplements can be adopted, which allow the nutritional requirements of the ruminal microbiota to be met initially and, subsequently, of the ruminants themselves [3]. The knowledge of the ruminal environment to allow conditions favorable to the growth of microorganisms is essential for the cattle to efficiently use the fibrous carbohydrates of the ingested fodder. Thus, to evaluate the effect of the microbial fermentation in rumen, it is necessary to measure the fermentation parameters and rumen microbial populations, which reflect the activity of the ruminal microbiota [4].

Great advances in ruminant nutrition in last three decades have basically occurred as a consequence of a better understanding of ruminal environment, particularly aspects related to nutrition, ecology, growth, and physiology of ruminal microbiota. In Brazil, research on rumen microbiology is emerging, but there are still few studies that relate dietary characteristics to ruminal microbiota, which restricts a clear understanding of interactions between ruminant animal, its diet, and ruminal microbial population in tropical conditions.

We tested the hypothesis that supplementation with protein improves fermentation parameters without damaging the rumen microbial populations of beef cattle grazing in dry season. This study evaluated the impacts of supplementation with concentrates with three levels of protein on intake, fermentation parameters, and rumen microbial population of Nellore bulls grazing Urochloa brizantha cv. Marandu during the dry season.

Materials and Methods

The protocol used in this experiment was in agreement with the Ethical Principles for Animal Research established by National Council for Control of Animal Experimentation (CONCEA). This experiment was approved by institutional Committee for Ethics in Use of Animals of the Federal University of Mato Grosso (Protocol Number 23108.060964/13-6), Cuiabá, Mato Grosso, Brazil.

Animals and Diet

The experiment was conducted at the Experimental Farm of Federal University of Mato Grosso (Cuiabá, Mato Grosso, Brazil), from April 2015 to July 2015, representing the dry season.

Four rumen-cannulated Nellore bulls, with an average age of 27 months and an initial body weight of 571 ± 31 kg, were used. The animals were allocated individually into 4 paddocks of 0.24 ha, consisting of Urochloa brizantha cv. Marandu. The animals were distributed in Latin square. The experimental period was 84 days, divided into four periods of 21 days each. The first 15 days of each period were destined to adapt the animals to the supplement, and the last six days were used for sampling.

The diets utilized consisted of Urochloa brizantha cv. Marandu pasture and concentrated supplement with three levels of protein: low, medium, and high protein supplements; the not supplemented treatment, in which animals receive only mineral salt, with the objective to evaluate rumen microbiome of animals not subjected to supplementation (Table 1).

Table 1 Composition of ingredients used in the supplement and chemical composition of supplements and Urochloa brizantha cv. Marandu (%DM basis)
Full size table

The low, medium, and high protein supplements were formulated to contain, respectively, 155, 515, and 875 g of CP/kg of dry matter and provided, respectively, 18, 60, and 100% of the protein requirements of Nellore bulls with an average body weight of 588 kg and an expected weight gain of 0.400 kg/animal/day [5]. The animals were supplemented at 1.5 kg/animal/day at 10 a.m.; water was provided ad libitum. The ingredients were sampled in each period.

Handling, Measurements, and Samples

The pastures grazed by bulls were sampled by clipping forage in average height (25 points in each paddock diagonal using a graduated stick) [6] in two sites (metal squares 0.25 m2) to a 5 cm residual sward height and the dried at 55 °C. The forage sample representing from each period was a composite of grass samples collected from each paddock grazed. Paddocks had an average forage mass of 3.24 ± 0.79 ton/ha of dry matter (DM) and an average sward height of 17.36 ± 3.8 cm. Furthermore, for nutrient composition, samples of forage were obtained by manual simulation of grazing [7].

Fecal DM excretion was determined by using titanium dioxide (15 g/animal/day) as an external marker that was packaged and introduced into rumen once daily at 09 a.m. for 8 days. Then, fecal grab samples were collected on 15th day (4 p.m.), 16th day (12 a.m.), and 17th day (7 a.m.) [8]. Fecal samples were air dried for 72 h (55 °C) and then ground in a knife mill to pass through a 1-mm screen sieve. Thus, fecal samples were composed and analyzed.

Ruminal digesta were collect at 01 p.m. (3 h after supplementation) on the 19th day of each period. It was sampled approximately 1.8 L of rumen contents (equal volume of liquid and solid) from three sites within the central and ventral portion of rumen (anterior, central, and posterior), which was immediately filtered through four layers of cheesecloth, where 25 mL of liquid and 25 g of solids were separately stored in falcon tube and immediately frozen − 20 °C, for subsequent DNA extraction. Prior to sampling, ruminal pH was measured using a digital potentiometer. An aliquot 50 mL of ruminal fluid added to 1 mL of sulfuric acid 1:1, conditioned in Falcon tube, identified and frozen at − 20 °C for analysis of ruminal ammonia nitrogen concentration (RAN), which was analyzed by distilling with 2 M KOH in a micro-Kjeldahl system, according to original procedures of Fenner [9].

Microbial DNA was extracted from the liquid phase along with the solid phase, but solid phase (25 g) was first homogenized in a chilled blender for 2 min with ice-cold extraction buffer (EB; 100 mMTris/HCl, 10 mM EDTA, 0.15 M NaCl pH 8.0) to release the solid-associated bacteria as described by [10]. After that, 25 mL of ruminal liquid was added to the centrifuge tubes of 250 mL, and all the materials will be centrifuged at 10,000×g/25 min/4 °C, 25 mL. The bacterial pellet was resuspended in 4 mL of ice-cold extraction buffer, and added 1000 µL to a 2.0 mL microtube with screw cap containing 0.5 g of glass beads (in duplicate). Subsequently, 50 µL of 20% sodium dodecyl sulfate (SDS) and 700 µL of phenol (pH 8.0) were added to microtubes. To lysis of microbial cells, microtubes were stirred in ball mill (in place of bead-beater), for 3 min, then heated in a 60 °C water bath for 10 min and stirred again in a ball mill for 3 min. After that, microtubes were spin in microfuge 18 g for 10 min and extraction was done with a combination of phenol/chloroform and precipitation with isopropanol [10]. The DNA obtained was resuspended in 100 µL ultrapure water. DNA concentrations were estimated by spectrophotometry (Thermo Scientific NanoDrop™ 1000) in 260 and 280 nm and the DNA preparations were stored at 4 °C for short-term use or were archived at − 20 °C.

“Regular” PCR was done on a thermocycler ProFlex PCR System (Life technologies®) in a 25-µL volume containing 10 ng de DNA, 1U Taq DNA polymerase (Invitrogen®), 0,2 mM dNTP’s, 0.5 mM MgCl2, 1 × buffer PCR 10 × (200 mM Tris–HCl pH 8, 4 500 mMKCl), and 20 pmol each primer. Reactions were performed under the following conditions: one cycle at 94 °C for 5 min, and 35 cycles of PCR (94 °C for 30 s, 57 °C for 35 s, and 72 °C for 30 s) with a final extension for 5 min at 72 °C. The PCR products were analyzed by running on 2% agarose gels containing GelRedTM (Nucleic Acid gel stain, Biotium®) and visualizing for a single specific band and the absence of primer dimer products in Chemi-Doc (Bio-RadTM).

For quantification of total bacteria and relative quantification of cellulolytic bacteria (Fibrobacter succinogenes, Ruminococcus albus, Ruminococcus flavefaciens, Butyrivibrio fibrisolvens), amylolytic bacteria (Selenomonas ruminantium), and Archaeas (Table 2), the technique used was qPCR. These species were chosen because they are the main cellulolytic and amylolytic strains.

Table 2 PCR primers used in this study
Full size table

The regular PCR described above was used to generate sample-derived DNA standards for each real-time PCR assay. Sets of such DNA standards were prepared from composite of microbiome DNA extracted from bulls. PCR products derived with each primer pair from each DNA set were pooled and purified using Kit GFXTM PCR DNA Purification Kit & Gel Band (GE Healthcare®). The DNA concentration was estimated with Qubit 2.0 instrument applying Qubit dsDNA HS Assay (Life Technologies, Invitrogen division, Darmstadt, Germany).

The copy number of each DNA standard was calculated based on mass concentration and average molecular weight of respective amplicons [13]. However, to standards of Selenomonas ruminantium, genomic DNA purified from a pure culture of target species. Tenfold serial dilution in Tris–EDTA of each DNA standard was prepared prior to real-time PCR assays. In total, five real-time PCR standards were prepared from set of microbiome DNA samples for three real-time PCR assays. Each of these standards was used in real-time PCR assays.

Each microAmp plate well contained the following: 2X SYBR Green Master Mix (Applied Biosystems), which contained all the nucleotides, polymerase, reaction buffer, and SYBR green dye; forward and reverse primers at concentrations of 300, 450, or 900 nM (depending on previously determined optimal concentration); and nuclease-free water to a total of 23 µL per well. To this solution, 2.0 µL of each sample or standard was added and plate was briefly centrifuged and placed in thermocycler for analysis.

Real-time PCR was performed using Applied Biosystems Step One sequence detection system with fluorescence detection of SYBR green dye. For all the primer pairs, amplification consisted of an initial hold for 10 min followed by 40 cycles of 95 °C for 15 s and 60–61 °C for 60 s depending on the Tm of the specific primer pair (Table 2).

Relative quantification was used to determine species proportion [14]. Calculations were provided for each rumen sample.

Chemical Composition Analysis

Forage and supplement ingredient samples were ground through a 1-mm screen of a Wiley mill, and were analyzed for dry matter, total nitrogen, ash according to AOAC [15]. Neutral detergent fiber, residual nitrogen compounds, and indigestible neutral detergent fiber (NDFi) were analyzed as described by Mertens [16], Licitra et al. [17], and Valente et al. [18], respectively. The estimation of the combined ether extract and non-fibrous carbohydrate contents of the substrates of supplements and forage was performed according to Hall [19].

Fecal DM excretion was estimated based on the ratio between the amount of marker supplied and its concentration in feces, according to the equation described by Holleman and White [20]: FE = TiO2 provided/TiO2 concentration feces.

The DM intake was estimated using the indigestible neutral detergent fiber as an internal marker and calculated by the equation proposed by Detmann et al. [21].

Statistical Analysis

The experiment was analyzed according to a 4 × 4 Latin square design model:

$${\text{Yijk}}=\mu +{\text{ti}}+{\text{aj}}+{\text{pk}}+{\text{eijk}},$$

where µ is general constant; ti is the treatment effect i; aj is the effect related to animal j; pk is effect for the experimental period k; and eijk is a random error, associated to each observation, which is assumed to be IIND (0; s2).

The data were analyzed using the mixed procedure of SAS (Statistical Analysis System, version 9.3). The variables were analyzed as a 4 × 4 Latin square design, to which the ANOVA was first employed to verify treatment effect, effect related to animal, and effect for the experimental period. Treatment was considered as fixed effect, and animal and period as random effect. Observing treatment effect was performed orthogonal contrast to specific partition of effects of treatment were compared: supplements (mineral supplement vs. low, medium, and high protein supplements); (low protein vs. medium protein) and (medium protein vs. high protein). All statistical procedures were performed using 0.05 as the critical level of probability.

Results

Bulls that received supplements had a greater (P < 0.05) crude protein (kg/day), EE + NFC (kg/day), and dry matter (% of body weight) intakes than animals on mineral salt (MS). However, there was no difference (P > 0.05) across the treatments for intakes of DM (kg/day), forage (kg/day), and NDF (kg/day and % of body weight) (Table 3). Furthermore, animals on medium protein supplement (MPS) had a higher (P < 0.05) intake of CP than animals on low protein supplement (LPSU) and had a lower (P < 0.05) CP intake of high protein supplements (HPS) (Table 3).

Table 3 Intake and digestibility of nutrients in beef cattle not supplemented and supplemented with concentrate with varying protein contents
Full size table

There was no difference (P > 0.05) across the treatments for digestibility of DM and NDF. However, bulls that received concentrate had a greater (P < 0.05) CP and EE + NFC digestibility when compared with animals on MS (Table 3). The MPS animals showed a higher (P < 0.05) digestibility of CP than the LPSU animals (Table 3).

The ruminal pH did not differ (P > 0.05) between the treatments evaluated (Table 4). However, animals supplemented had greater (P < 0.05) concentrations of ruminal ammonia nitrogen (RAN) than the animals received only mineral salt (Table 3).

Table 4 pH and ruminal ammonia nitrogen data of bulls grazing tropical forage
Full size table

The concentrations of RAN in animals of medium protein supplement (MPS) were greater (P < 0.05) when compared with low protein supplement (LPSU) and were lower (P < 0.05) in relation to animals of high protein supplement.

The bulls that received concentrate had greater (P < 0.05) proportions of Butyrivibrio fibrisolvens, lower (P < 0.05) proportions of Fibrobacter succinogenes, Ruminococcus albus, Ruminococcus flavefaciens, and Methanogen Archaea than bulls of mineral salt (Table 5).

Table 5 Percentages of target species in ruminal samples relative to total Bacteria content
Full size table

Animals of low protein supplement had greater (P < 0.05) proportions of Fibrobacter succinogenes, Ruminococcus flavefaciens, and Selenomonas ruminantium than those received medium protein supplement. However, no differences (P > 0.05) in terms of the proportions of Ruminococcus albus, Butyrivibrio fibrisolvens, and Methanogen Archaea were observed between animals supplemented with low and medium protein supplements (Table 5).

Furthermore, animals fed with high protein supplement had greater (P < 0.05) proportion of Ruminococcus albus when compared with MPS. However, it was observed that these supplements did not promote differences (P > 0.05) on proportions of Fibrobacter succinogenes, Ruminococcus flavefaciens, Selenomonas ruminantium, Butyrivibrio fibrisolvens, and Methanogen Archaea.

The Methanogen Archaea had a greater (P < 0.05) proportion in ruminal samples of animals supplemented with the levels of protein in relation to the animals which received only mineral salt (Table 4). After Archaea, the most common species were Ruminococcus flavefaciens and Selenomonas ruminantium, each of whose 16S RNA gene copies generally accounted for 0.08–0.22% of total bacteria. Fibrobacter succinogenes, Ruminococcus albus, and Butyrivibrio fibrisolvens were species less abundant, each generally in 0.02–0.07% of bacterial 16S rRNA gene copies. In total, 6 individual species accounted for only about 3.8 to 5.3% of bacterial 16S rRNA gene copies in these ruminal samples.

Discussion

It was observed that 70% of crude protein content on the forage is associated with cell wall, in the form of neutral detergent insoluble protein, which is slowly degraded in rumen or unavailable. Nevertheless, this did not affect the forage, dry matter, and neutral detergent fiber intakes, which reflects the fact that the forage probably had no deficiency of nitrogen for the growth of fibrolytic bacterias [22].

Furthermore, the greater intake and digestibility of CP observed in animals supplemented in relation to MS animals may be due to greater CP concentrations in the supplement in relation to the forage [23]. Furthermore, the greater intake and digestibility of CP in MPS animals in relation to LPSU, and the greater intake of CP in HPS animals when compared with MPS animals, can be explained with the lower protein content in LPSU, while HPS had relatively high protein content.

Already the less amount of EE + NFC in the forage resulted in lower intake and digestibility of EE + NFC in animals fed exclusively on pasture, once this nutrients presents in the forage was less than the 35% suggested by Valadares et al. [24] as the ideal level to maximize the use of dietary non-protein nitrogen compounds, since is readily usable substrate by rumen microorganisms.

The supplement protein and the lower EE + NFC of the forage were not able to alter ruminal pH, since, the ruminal pH (6.35–6.70) was within the range considered acceptable for fiber digestion, since below 6.2, the growth of the cellulolytic bacteria population is retarded [25] or reduced and promotes growth of amylolytic bacteria, consequently limiting the digestion of fibers [26].

However, the increased CP levels in the supplements and increased CP ruminal digestibility are related with the increase in RAN concentration, which means an overall improvement in nitrogen availability in the animal gastrointestinal tract and increased animal metabolism [27].

Thus, the concentration of RAN in the rumen results from the balance between production, absorption, and incorporation of nitrogen into microbial protein [28]. The greater RAN concentrations (16.38 mg/dL) observed in animals supplemented are due to the fact that the concentrate had higher amounts of protein (Table 1) when compared with mineral supplement, which means an improvement in nitrogen availability in the animal gastrointestinal tract and also in the animal metabolism [27]. In addition, the RAN concentration in animals from mineral salt (6.55 mg/dL) was below 8.4 mg/dL as reported by Detmann et al. [29] for an equilibrium between the inflow and outflow of nitrogen in the rumen of animals fed tropical forage, which resulted in lower nitrogen balance in the rumen values.

Furthermore, the decreased RAN concentrations observed in medium protein-supplemented animals in relation to high protein supplements can be due to the non-fibrous carbohydrate which imply a better microbial assimilation [22].

Therefore, the supply of non-fibrous carbohydrate favored a reduction in proportions of Fibrobacter succinogenes, Ruminococcus albus and Ruminococcus flavefaciens in supplemented animals when compared with animals that received only mineral salt, since, these bacteria are considered the mainly species of cellulolytic bacteria [30] and by their trophic niches be cellular wall and pure cellulose [31]. In addition, the dependency of ammonia to growth of Ruminococcus albus bacteria [32] proportioned the increased this specie in rumen of animal with high protein supplementation.

However, the possible increase of branched-chain volatile fatty acids (BCVFA) in low protein supplement favored the increase in proportions of Fibrobacter succinogenes, Ruminococcus flavefaciens and Selenomonas ruminantium in relation to medium protein supplement, since ruminal cellulolytic bacteria and some amylolytic bacteria require BCVFA (isovaleric, isobutyric, and 2-methylbutyric) as well as nitrogen for growth [33, 34, 35]. Furthermore, we can observe that the forage provided nitrogen enough to growth these bacteria species and the RAN was better utilized, since lower concentrations of RAN were available in rumen.

Furthermore, the supply of concentrate to animals favored increase of 53% of Butyrivibrio fibrisolvens in relation to animals that received only mineral salt. This result can be due to the bacteria utilizing the starch, diverse sugar, beyond cellular wall, pure cellulose, hemicellulose and pectin [31]. Furthermore, according to Cotta and Hespell [36], the strain Butyrivibrio fibrisolvens HI7c has higher proteolytic activity.

However no was expected the greater proportions of Methanogen archaea in animals supplemented, since these rumen microorganisms utilize H2 and CO2 produced by the protozoa, fungi and bacteria from the catabolism of hexoses to produce CH4 and generate ATP [37, 38], which benefits the animals by utilization of electron for reducing equivalents to minimize the partial pressure of H2 [39] inside the rumen. In the rumen, H2 is produced during plant cell wall degradation as an intermediate compound by cellulolytic bacteria and anaerobic fungi [40]. On the other hand, the archaea can be associated with protozoa (not analyzed in this study) in animals supplemented, since this association account for approximately 37% of the methane emitted by ruminants [41].

In addition, most of the archaea found in the rumen belong to the genera: Methanobrevibacter, Methanosarcina, Methanomicrobium, and Methanobacterium [42].

Each methanogenic species has a substrate preference, and most can use only one or two substrates [43]. In addition, the differences in methanogenic diversity are attributed to dietary variation [44] or to dietary and host characteristics [45, 46].

Thus, the most abundant cellulolytic specie was Ruminococcus flavefaciens, in which the 16S RNA gene copies accounted for 0.12–0.22% of the total bacteria. This result contradicts the Michalet-Doreau et al. [47] that reported that Fibrobacter succinogenes is the rumen-dominant population in sheep that was fed alfalfa hay diet. Thus, the quantification of ruminal microorganisms is essential for the design of strategies to produce meat to pasture efficiently. Further studies are needed in order to allow a deeper understanding of the ruminal microbiota of animals grazing tropical forage.

Conclusion

The supply of 515 g/day protein via supplement for grazing beef cattle during dry season provides better fermentation parameters and improves ruminal conditions for the growth of cellulolytic bacteria.