Effects of urea plus nitrate pretreated rice straw and corn oil supplementation on fiber digestibility, nitrogen balance, rumen fermentation, microbiota and methane emissions in goats

Urea plus nitrate pretreated rice straw

The rice straw originated from Jiangxi province of China. Whole plant material was obtained and thrashed to remove the grain and the straw was sun-cured and chopped into pieces of approximately 3 cm in length. Straw was pretreated with urea (34 g/kg straw dry matter [DM]) and 6 g/kg DM of ammonium nitrate (to supply 4.7 nitrate g/kg straw DM) for the urea plus nitrate treatment (UN). The pretreatment procedure was fully described by Zhang et al. [9]. Briefly, urea and ammonium nitrate were dissolved in water (40%, w/v) and sprayed onto the straw to achieve the appropriate concentrations. The pretreated straw was then placed into sealed 200 L plastic vessels and incubated at 15 ± 3.0 °C (means ± SD) for 4 weeks. After incubation, the straw was sun-cured for feeding.

Experimental design, goats, and diets

Nine female Liuyang Black goats (a local breed in southern China, 1 year of age) with body weight of 19.0 ± 1.22 kg (mean ± SD) at the start of the experiment were used. The experiment was a triple 3 × 3 Latin Square design with 3 dietary treatments (Table 1). The control diet was formulated to meet appropriately 1.2 to 1.3 times the digestible energy (DE) and CP requirements of non-pregnant female goats according to Zhang and Zhang [12], and consisted of 50% rice straw and 50% concentrate (Table 1). The UN diet was formulated by replacing untreated rice straw with UN pretreated rice straw. The combination of UN and corn oil supplementation (UNCO) diet was formulated by replacing soybean meal (15 g/kg of dietary DM) and corn grain (15 g/kg of dietary DM) with corn oil (30 g/kg of dietary DM). Goats were fed individually at 07:30 and 17:30 h with equal portions of feed offered at each meal.

The experiment consisted of 3 periods, each with 27 d including 14 d for diet adaptation, 5 d for collecting feces and urine, 6 d for measuring CH₄ emissions and 2 d for rumen sampling. All goats were housed in individual metabolism cages and had free access to drinking water.

Apparent total-tract digestibility and N balance

Total feces and urine were collected and weighed twice daily from d 15 to 19. A subsample (~ 1%) of feces and urine from each animal was obtained at each collection time and frozen immediately at − 20 °C, and another subsample (~ 1%) was acidified using 10% (w/w) H₂SO₄ to prevent N loss and then frozen immediately at − 20 °C. The subsamples were then individually combined by day and goat within period. The acidified samples were used for total N analysis, whereas non-acidified samples were used for other chemical analysis.

The N balance, including N excretion and retention, was calculated using the following equations:

N excretion (g/d) = fecal N (g/d) + urinary N (g/d).

N excretion (%) = [N excretion (g/d) / N intake (g/d)] × 100%.

N retention (g/d) = N intake (g/d) – N excretion (g/d).

N retention (%) = [N retention (g/d) / N intake (g/d)] × 100%.

Enteric methane emission

After total collection, CH₄ emissions were measured for 2 d per goat by moving each goat into a respiration chamber as described by Wang et al. [13]. As only three chambers were available, it took 6 d (from d 20 to 25) to complete the measurements for all goats. Nine goats were assigned to 3 blocks, with each block containing 3 goats (1 goat from each treatment). The goats in each block were assigned to the same chamber and 1 goat from each block was used for each of the 2-day sequential measurements. The goats were restrained but had free access to a feed bin and drinking water within the chamber. On average, the flow rate of air was maintained at 40 m³/h. The CH₄ concentration in outlet gas from each chamber and ambient gas were measured using an Ultraportable Greenhouse Gas Analyzer (MIU-374-8, Los Gatos Research, San Jose, CA 95134, USA). The cycling time to separately measure CH₄ concentration from the 3 chambers was 30 min, with 8 min for each chamber gas and 6 min for ambient gas. Calibrations were done by discharging purified CH₄ into each chamber at 25 cm³/min controlled by a gas flow meter (C100L-CRWE-DD, SIYA Detection Instrument Co., Ltd., Shanghai, China). The gas flow data from both the gas flow meter and chamber were calculated and compared, and then the data for the chamber were adjusted to 100% recovery according to the data of the gas flow meter. The chambers were opened twice a day at 07:30 and 17:30 h to deliver the diets, and the chambers were cleaned before the morning feeding.

Rumen sampling

Rumen contents were collected at 0, 2.5 and 6 h after morning feeding on 2 consecutive days (d 26 and d 27) using an oral stomach tube according to Wang et al. [13]. About 150 mL of rumen contents were rapidly collected after discarding the initial 100 mL of rumen contents. Two subsamples (15 mL each) were immediately frozen at − 80 °C in liquid nitrogen for DNA extraction and subsequent microbial analysis. Two other subsamples (35 mL each) were immediately transferred into 50-mL plastic syringes for measuring dissolved hydrogen (dH₂) and dissolved CH₄ (dCH₄) concentrations according to the procedures outlined by Wang et al. [14]. About 10 mL of rumen contents were used for the measurement of ruminal pH using a portable pH meter (Starter 300; Ohaus Instruments Co. Ltd., Shanghai, China). An aliquot of 2.5 mL of rumen contents was centrifuged at 15,000×g for 10 min at 4 °C, and supernatant (1.5 mL) was transferred into tubes containing 0.15 mL of 25% (w/v) metaphosphoric acid, vigorously hand-shaken and stored at − 20 °C for subsequent determination of volatile fatty acid (VFA) and ammonia concentration.

Sample analyses

The samples of feed, orts and feces were dried in a forced-air oven at 65 °C for 48 h, ground through a 0.25-mm screen (WJX-100, Shanghai, Yuanwo Company) and analyzed in triplicate for DM, OM, ether extract (EE), neutral detergent fiber (NDF), acid detergent fiber (ADF), N content and gross energy (GE). The DM (method 930.15), OM (method 942.05), EE (method 963.15) and N (method 970.22) were analyzed according to published methodologies [15]. Neutral detergent fiber and ADF were assayed according to the method of Van Soest et al. [16], and expressed inclusive of residual ash. Heat stable α-amylase was added during the NDF analysis. Gross energy was determined using an isothermal automatic calorimeter (5E-AC8018, Changsha Kaiyuan Instruments Co., Ltd., China). Individual VFA concentrations were measured using a gas chromatograph (Agilent 7890, Palo Alto, CA, USA) according to the procedure described by Wang et al. [14]. Ammonia-N concentration was measured by UV–Vis spectrophotometer (Shimadzu UV-2450, Kyoto, Japan) according to the protocol described by Chaney and Marbach [17].

Microbial analyses

Rumen samples taken 2.5 h after the morning feeding were freeze-dried (CHRIST RVC2–25 CDPIUS, Marin Christ CO., Ltd., Osterode, Germany) for microbial analysis. The DNA was extracted with repeated bead beating plus column purification as described by Yu and Morrison [18], and eluted by 300 μL TE buffer (Tris 10 mmol/L, EDTA 1 mmol/L, pH = 8.0). The quality and quantity of DNA were measured based on absorbance at 260 and 280 nm using a NanoDrop ND-2000 (NanoDrop Technologies Inc., Wilmington, USA).

The 16S rRNA gene V3-V4 hypervariable regions of bacteria genomic DNA were used for PCR amplification with the primers 5′- ACTCCTACGGGAGGCAGCAG-3´ (338F) and 5′- GGACTACHVGGGTWTCTAAT-3′ (806R). The PCR reactions were performed in triplicate using a 20-μL mixture containing 0.8 μL of each primer, 10 ng of template DNA, 2 μL 2.5 mmol/L dNTPs, 0.4 μL of FastPfu polymerase (Transgen, Beijing, China) and 4 μL 5 × FastPfu Buffer (Transgen, Beijing, China). The thermal cycling programing was performed as follows: initial denaturation step, 95 °C, 3 min; denaturation, 27 cycles, 95 °C, 30 s; annealing, 55 °C, 30 s; elongation, 72 °C, 45 s; and final extension, 72 °C, 10 min. The PCR products were excised from 2% agarose gels and purified using a QIAquick Gel extraction kit (Qiagen, Hilden, Germany). Amplicons from each reaction mixture were quantified fluorometrically, normalized and pooled at equimolar ratios based on the concentration of each amplicon. Amplicons were sequenced with the Illumina MiSeq platform (PE300, Majorbio Bio-Pham Technology, Shanghai, China). Quality control of the sequence reads was performed using MOTHUR v.1.39.5 [19] following the protocol described by Kozich et al. [20]. The high-quality reads were clustered into operational taxonomic units (OTU) at 97% similarity using Usearch v.7.0 [21]. Representative sequences defined by abundance from each OTU were identified using PyNAST [22] against the SILVA database for bacteria [23]. Taxonomy analysis was performed using the Ribosomal Database Project classifier v.11.1 [24] with a minimum support threshold of 80%. Alpha diversity analyses were generated, including observed species (Sobs), Chao, Ace, Shannon-Weiner, and Simpson’s indices. All 16S rRNA gene sequences were deposited into the NCBI Sequence Read Archive under accession number SRP155601.

Real-time quantitative PCR (qPCR) was performed according to the procedures described by Jiao et al. [25]. Briefly, the plasmid DNA containing exact 16S or18S rRNA gene inserts were used to make a standard curve for selected microbial groups. All standard curves met the following requirements (R² > 0.99, 90% < E < 120%). The protozoa, fungi, total bacteria, total methanogens, and four selected bacteria were studied. Fibrobacter succinogenes (F. succinogenes), Ruminococcus albus (R. albus) and Ruminococcus flavefaciens (R. flavefaciens) were selected because of their predominant role in fiber digestion [26]; and Selenomonas ruminantium (S. ruminantium) was selected because it plays a role in nitrate and nitrite reduction [27]. Forward and reverse primers of the selected microbial groups are shown in Table 2.

Statistical analyses

The statistical analysis was performed using the mixed linear model procedure of SPSS 19.0 (Chicago, IL, USA). Fermentation and CH₄ emissions data were averaged to obtain a mean value for the 2 consecutive days before further analysis. For the data of CH₄ emissions, digestibility, qPCR, relative abundance and alpha diversity indexes estimated from 16S rRNA gene library sequences, the model used included treatment (n = 3) as a fixed effect, and period (n = 3) and animal (n = 9) as random effects. When sampling time was included, the model included dietary treatment (n = 3) and the interaction of treatment and sampling time as a fixed effect, sampling time (n = 3) as a repeated measurement, and animal (n = 9) and period (n = 3) as random effects. When significant differences were found, a multiple comparison was conducted to elucidate differences between two particular treatments, and P-values were adjusted using the Bonferroni method. Statistical significance was declared at P ≤ 0.05, with a tendency towards significance declared at 0.05 < P ≤ 0.10.

Feed digestibility

It has been reported that UN pretreatment improved ruminal degradation of rice straw measured in in vitro batch incubation [9]. The present in vivo study also indicates that goats fed UN had greater total-tract digestibility of NDF than those fed control. Urea in the UN pretreatment can be converted to ammonia during incubation with rice straw, which removes the polymerized silica-waxy compounds from the leaf fractions and destroys the covalent association between lignocellulose and exposes the inner tissues to bacterial colonization [28]. Increased in vivo fiber digestibility indicates that UN pretreatment allows greater access of ruminal fibrolytic bacteria to the fiber substrate. Griffith et al. [29] reported that ammoniation of fiber facilitates its degradation by rumen microbes. The rumen has a wide-range of fibrolytic microbial groups that utilize the fibrous substrates in feed. Protozoa and fungi are active fiber degraders, while other predominant fibrolytic bacteria include R. albus, R. flavefaciens and F. succinogenes [26]. In our study, UN pretreatment increased protozoa and R. albus populations, which may have contributed to increased total-tract digestibility of NDF in goats.

Although supplementation of corn oil to the UN diet did not further improve nutrient digestibility, fiber digestibility remained greater than that of the control treatment. It has been reported that high concentrations (> 40 g fat/kg DM) of fat supplementation exert negative effects on feed digestibility [30]. Thus, the relatively low dose of corn oil (30 g/kg DM) supplementation was not expected to decrease fiber digestibility. Similarly, Machmüller and Kreuzer [31] reported that 35 g oil /kg DM supplementation did not alter fiber digestibility in sheep. Although protozoa are major fiber degraders in the rumen [26], the decreased protozoa population observed for UNCO did not negatively affect fiber digestibility.

Nitrogen balance

The greater N intake of goats fed UN did not alter total-tract digestibility of CP, but it increased rumen ammonia-N concentration and urinary N excretion, leading to a decrease in N retention. Djajanegara and Doyle [32] reported that urea pretreatment of rice straw increased fecal N excretion in sheep. Gunun et al. [5] reported that urea pretreated rice straw increased fecal and urinary N excretion in dairy cows. The urea and nitrate in the UN pretreatment provided additional non-protein N for rumen microbes. The lack of treatment effects on 16S rRNA gene copies of bacteria would suggest that the additional N supplied did not promote bacterial growth. It is likely that the soluble protein in the control diet was adequate for microbial growth, and additional N supplementation did not exert further benefits to the host animals. Furthermore, the greater observed protozoa population for the goats fed UN may have contributed to the decreased capture of ruminal N, resulting in increased ruminal ammonia concentration and urinary N excretion, because protozoa can prey on bacteria and reduce the amount of N used for bacterial protein synthesis [10].

Corn oil supplementation to the UN diet decreased N excretion and increased N retention. Benchaar et al. [33] reported that 40 g oil/kg DM reduced N excretion in dairy cows. Doranalli and Mutsvangwa [34] found that 60 g oil/kg DM decreased total N excretion and increased N retention in growing lambs. Improved N retention might be due to decreased intra-ruminal N recycling and increased microbial N flow to the duodenum [10]. In our study, corn oil supplementation to the UN diet decreased ruminal protozoa numbers, which has been reported to decrease predation of bacteria, and thus facilitate N capture by bacteria [10]. Therefore, supplementing corn oil to the UN diet offset the negative impact of UN pretreatment on N utilization, leading to an improvement of N retention in goats.

Methane emission and rumen environment

Enteric CH₄ emission was on average 9.9 ± 1.35 g/d, which is less than in other studies using goats or sheep. This discrepancy is likely due to the lower body weight (19.0 ± 1.22 kg) of goats used in the present study. Machmüller and Kreuzer [31] reported 29.4 g/d CH₄ emissions in sheep with body weight of 78 ± 7.0 kg, while Abecia et al. [35] observed 15.9 g/d CH₄ emissions in goats with average body weight of 43 ± 1.7 kg.

The UN treatment greatly decreased CH₄ emissions in terms of g/kg DMI and g/kg OM digested, but did not alter daily CH₄ emission in terms of g/d. This effect is due to the slight increase in DMI and OM digestibility in goats fed the UN diet. Nitrate acts as an alternative hydrogen sink and inhibits CH₄ production in the rumen. When compared with control, the observed increase in protozoa and S. ruminantium populations for UN is consistent with the roles of these microorganisms in nitrate and nitrite reduction [27, 36]. Stoichiometrically, 4 mol of 2[H] can be redirected to ammonia away from methanogenesis during 1 mol of nitrate reduction to ammonia (NO₃⁻ + 4 [2H] + 2H⁺ → NH₄⁺ + 3H₂O), equivalent to 258.1 g CH₄/kg nitrate. If 2.33 g nitrate/kg dietary DM in UN was completely reduced to ammonia, CH₄ reduction would have been 0.6 g/kg DMI. However, the observed CH₄ reduction of 2.1 g/ kg DMI is far more than that theoretical reduction. Toxic effects of nitrate and nitrite have been reported for methanogens, but the dietary nitrate content (2.33 g/kg DM) in the present study is far less than that in published studies (6.7 and 5.3 g/kg DM, respectively) [7, 27], indicating that toxic effects may not have been the main cause of CH₄ reduction for the UN treatment.

The shift of rumen fermentation may contribute to the CH₄ reduction for the UN treatment. The decreased butyrate molar proportion and (acetate + butyrate) to propionate ratio of the goats fed UN is consistent with the decrease in CH₄ emission of UN compared with control. Acetate and butyrate formation results in net [H] production, while propionate formation incorporates [H] [37]. A shortage of [H] would result in reduced methanogenesis. Our previous in vitro study found that urea plus nitrate pretreatment for rice straw decreased CH₄ production with a reduction of acetate to propionate ratio [9]. van Zijderveld et al. [38] observed that nitrate supplementation caused CH₄ inhibition, which was accompanied by increased propionate molar proportion at the expense of acetate and butyrate.

Oil supplementation inhibits CH₄ production [39], due to a reduction in OM intake, rumen fermentation and protozoa associated methanogens. Guyader et al. [40] reported that a combination of 40 g linseed oil/kg dietary DM and 30 g calcium nitrate/kg dietary DM shows additive effects on CH₄ mitigation. However, we did not find additive effects on CH₄ mitigation when corn oil was supplemented to the UN diet. Therefore, the combination of UN and corn oil supplementation effectively decreases CH₄ emissions, but not to a greater extent than UN alone.

As nitrate acts as an electron acceptor in anaerobic environments, reduction of nitrate to ammonia is associated with decreased H₂ availability for methanogens [41]. However, nitrate or nitrite may exert toxic effect on methanogens, leading to an increase in ruminal H₂ accumulation or dH₂ concentration [39]. The balance between nitrate reduction and nitrite toxicity determines the available H₂ in the rumen. In our study, UN treatment decreased ruminal dH₂ concentration, indicating that nitrate in UN treatment incorporates [H] for ammonia synthesis during nitrate reduction. The low nitrate content (2.33 g/kg dietary DM) of the UN treatment was unlikely to have been toxic to the methanogens. Urea plus nitrate treatment did not alter dCH₄ concentration, possibly because increased DMI and NDF digestibility can cause greater CH₄ synthesis. When compared with control, goats fed UNCO had lower dCH₄ concentration in rumen fluid, which is consistent with decreased CH₄ emissions. However, in comparison with control, UNCO increased the fiber digestibility and ruminal ammonia concentration, but did not alter major individual VFA profile. These results are surprising and the underlying mechanism needs further investigation.

Bacterial community

Urea plus nitrate treatment did not alter diversity of bacterial community, as shown by the similar Shannon-Weiner and Simpson indexes between UN and control groups. Bacteroidetes and Firmicutes are the two predominant phylums in the rumen bacterial community [42]. Although UN did not alter abundances of these phyla in comparison with control, the large increase in abundance of Ruminococcaceae family, which plays a very important role in fiber degradation [43], is consistent with the observed increase in fiber digestibility [44]. The greater 16S rRNA gene copies of R. albus of goats fed UN compared with control goats is consistent with the literature. Nguyen et al. [4] fed swamp buffaloes rice straw pretreated with urea plus limestone (20 + 20 g/kg DM of rice straw) and reported an increase in 16S rRNA gene copies of R. albus. It seems that the urea component in the UN treatment destroys the indigestible structural polysaccharide matrix, which facilitates colonization of fibrolytic bacteria by providing attachment sites for adherence of the bacteria. When corn oil was supplemented to the UN diet, abundance of Ruminococcaceae family was reduced, indicating that the unsaturated fatty acids in the oil had negative effects on microbiota of the rumen [45]. High-fat diets can decrease the abundance of Ruminococcaceae family [46]. When compared with control, UNCO did not alter bacterial diversity and abundance of major dominant bacteria and 16S rRNA gene copies of selected bacteria. Thus, it appears that corn oil supplementation can reverse the negative effects of UN on bacteria, resulting in a similar bacterial community as the control group.