Animal ethics approval (ACUC-KKU 32/61) was issued to ensure standard care of animals during the study.
Experimental design and treatments
A 3×2×4 in a completely randomized design were conducted. Factor A was level of sulfur at 0%, 1%, and 2% of concentrate dry matter (DM), factor B was level of urea at 2% and 4% of concentrate DM, and factor C was level of the FCR at 0, 200, 300, and 400 mg of the total substrate. The FCR (Manihot esculenta Kasetsart 50) at one-year-old of age was purchased from a local supplier located in Khon Kaen province, Thailand. Sulfur and urea were purchased commercially.
Substrate preparation
The substrates including rice straw and concentrate mixture were dried at 60°C and ground to pass a 1-mm sieve (Cyclotech Mill, Tecator, Sweden), while FCR was used as a fresh form. The ground samples of FCR, rice straw, and concentrate mixture were used to analyze DM (ID 967.03), organic matter (OM, ID 942.05), and crude protein (CP, ID 984.13) using the method of AOAC [24], neutral detergent fiber (NDF) and acid detergent fiber (ADF) according to Van Soest et al. [25]. Content of HCN in FCR was analyzed by using spectrophotometry (SpectroSC, LaboMed, inc, USA) with the 2,4-quinolinediol-pyridine reagent [26]. The concentrate ingredients and chemical compositions of concentrate, rice straw, and FCR used in this study were provided in Table 1.
Animals and rumen fluid provision
Two male rumen-fistulated dairy steers with body weight (BW) of 400 ± 50 kg were raised in a separate pen with accessible clean water and fed concentrate at 0.5% BW/day. The concentrate was formulated to have 12% CP following the recommendation of NRC [27]. Rice straw was daily fed ad libitum. The feeding lasted for 14-days before ruminal fluid was collected. After 14-days of feeding, approximately 1500 mL of ruminal fluid were manually collected and filtered through cheesecloth (four-layers) into pre-warmed thermos flasks, then immediately transferred to the laboratory.
Inoculum preparation and in vitro fermentation
The inoculum was made of the ruminal fluid and artificial saliva. The artificial saliva was prepared according to Menke and Steingass [28]. A 1:2 ratio of ruminal fluid and artificial saliva was mixed in a thermos flask to form the inoculum, warmed at 39°C, and continuously supplied with carbon dioxide. A 369 serum bottles (150 ml volume) were prepared, in which 72 serum bottles with 3 bottles for blank were used to study the kinetics of gas, 147 bottles used to study ruminal fermentation (pH, ammonia nitrogen-NH3-N, volatile fatty acid- VFA, and protozoa) at 4 and 6 h of incubation, and 147 bottles used to study the degradability at 12 and 24 h of incubation. All treatments were done in three replications. The ground concentrate mixture and rice straw were weighed into the serum bottles at 50:50 ratio to obtain the final substrate of 500 mg. The ground FCR (fresh form) was weighed into the bottles at its respective levels of total substrate. A 50 ml of artificial inoculum was withdrawn and injected into the serum bottles containing their respective treatments’ substrate. The bottles were then transferred to the water bath with pre-set temperature of 39°C and incubated at various time series.
Sample collection and analysis
The gas produced from fermentation was manually measured using a pressure transducer syringe at 0, 0.5, 1, 2, 4, 6, 8, 12, 18, 24, 48, 72, and 96 h of incubation. The amount of gas at each time of incubation was fitted to the gas equation of Ørskov and McDonald [29] to study the kinetics of gas as follows:
y = a + b[1 − e(−ct)]
where a is the gas production from the immediately soluble
fraction, b is the gas production from the insoluble fraction, c the gas production rate constant for the insoluble fraction (b), a + b is the potential extent of gas production, and t the incubation time.
After incubated for 4 and 6 h, the pH was measured using a Hanna pH meter (model HI83141, HANA instruments, Romania) from 147 bottles, and the liquid samples were then filtered through cheesecloth (four-layers) and centrifuged at 16,000× g for 15 min. After centrifuged, the supernatant was collected by dividing into two parts: the first part was used to analyzed NH3-N concentration using Kjeldahl methods according to AOAC [24] and VFA proportions including acetate (C2), propionate (C3), and butyrate (C4) using high-performance liquid chromatography (Instruments by controller water model 600E, Water model 484 UV detector, column Novapak C18, column size 4 × 150 mm, mobile phase 10 mM H2PO4 (pH 2.5); ETL Testing Laboratory, Inc., Cortland, NY). The remaining part was mixed with formaldehyde at 1:9 ratio for protozoal counts using microscopic (Boeco, Hamburg, Germany). HCN concentration in the liquid samples was measured by using spectrophotometry [26].
After incubated for 12 h and 24 h, the samples were collected by filtering through pre-weighed Gooch crucibles, then the Gooch crucibles containing sample were oven-dried at 60°C for 24 h. After oven-dried, the DM of samples and blank was used to calculate the in vitro DM degradability (IVDMD) [30]. Then, the samples were analyzed for in vitro NDF, and ADF degradability according to Van Soest et al. [25].
Statistical analysis
All data were subjected to the General Linear Models (GLM) procedures of SAS [31]. The following model was used:
yijkl = µ + ai+ bj+ ck + abij + acik + bcjk +abcijk + ɛijkl
where y is the observation, m is the overall mean, ai is the level of sulfur(i,1–3), bj is the level of urea (j, 1–2), ck is the level of FCR at 0%, 40%, 60% and 80% of all diet (k,1–4), abij, acik, bcjk, abcijk, is the interaction effect and ɛijkl is the error. Differences among treatment means for all parameters were contrasted by Tukey’s Multiple Comparison Test. Differences among means were accepted at P < 0.05.