Animals, treatments, and management
This study was started in July 2018 on a commercial dairy farm located in the City of Dongying, Shandong Province, China. The Institutional Animal Care and Use Committee of the Institute of Animal Sciences at the Chinese Academy of Agricultural Sciences approved all the experimental procedures (protocol no. IAS 20180115). Forty healthy female Holstein calves (4-day-old, 40 ± 5 kg of BW) born within one week on that dairy farm were recruited and separated from their dam immediately after birth. They were placed in individual south-facing Calf-Tel hutches (Hampel Corp., Germantown, WI) approximately 1.5 m apart. The hutches were bedded with sand and placed on a sand base. The calves were randomly allocated to 1 of 4 treatment groups (n = 10 calves per group) using the RAND function of Excel. The control group was fed no SB (SB0), while three treatment groups were fed SB (98.5% purity, Enkefu Co. Ltd., Beijing) mixed into liquid feeds (milk and then MR) at 15 (SB15), 30 (SB30), or 45 (SB45) grams per day per calf. The doses of SB supplementation were based on the study of Slusarczyk et al.  who fed calves SB at 2.2, 7.3, or 22 grams per day before weaning.
Prior to the feeding experiment, all calves were fed 4 L of colostrum within 1 h after birth and then two more feedings of colostrum 6 h (2 L) and 18 h (1 L) later. All calves were fed per the feeding regimen of the dairy farm. Specifically, the calves were fed only milk from 2 to 20 days of age. From 21 to 23 days of age, the calves were fed a mixture of milk and MR (Eurolac Blue, Netherlands) at different milk:MR volumetric ratio: 75% milk and 25% MR at 21 days of age, 50% milk and 50% MR at 22 days of age, 25% milk and 75% MR at 23 days of age. All the calves were fed only MR from 24 to 60 days of age (end of the experiment). The MR was dissolved in water to a final total solid content of 17.86%. All calves were fed the liquid feeds (milk, MR, and the mixture of both) using individual open buckets twice daily at 07:00 h and 15:00 h according to the following feeding regimen: 2.5 L/meal from 2 to 7 days of age, 3 L/meal from 8 to 10 days of age, 3.5 L/meal at 11 days of age, 4 L/meal at 12 days of age, 4.5 L/meal from 13 to 30 days of age, 6.5 L/meal from 31 days to 50 days of age, 5.5 L/meal at 51 days of age, 4.5 L/meal at 52 days of age, and then the allowance of the previous day with 1 L decrement per day (0.5 L/meal) until weaning at 60 days of age. The preset amounts of SB and liquid feed allowance were added together to feeding buckets and manually stir-mixed to dissolve the SB prior to each feeding. Each of the daily doses of SB was divided into 2 equal portions and fed in the morning and the afternoon. A pelleted starter feed (≥ 24.0% CP declared by the manufacturer, Rubeiyou8100, Yuan Xing Co., Ltd., China) was offered to the calves once daily after the liquid feeding in the morning from 4 days of age onward. When the starter ort was less than 20 g, an additional 100 g of starter was added the next day to ensure adequate starter was available all the time. The chemical composition of the experimental feeds (milk, MR, and starter) is presented in Table 1.
Calf growth measurement, sample collection, and analysis
Body length, BW, wither height, and heart girth were recorded at the beginning, 4, 14, 28, 42, and 60 days of age before the morning feeding. Average daily gain was calculated over the above time intervals and the entire experiment period. Intake of starter feed was recorded daily and for each calf at 09:00 h, and intake of liquid feed was recorded twice daily for each calf. Total dry matter intake (DMI) was calculated based on the consumption of the liquid feed and starter for each calf. Feed-to-gain (F:G) ratio was calculated as the ratio of total DMI to ADG.
Rumen fluid (about 25 mL) was collected from each calf two hours after the morning feeding of the liquid feed at 14, 28, and 60 days of age via a flexible esophageal tube (2 mm wall thickness, 6 mm i.d.) and a pump (Anscitech Co. Ltd., Wuhan, Hubei, China). The first 5 mL was discarded to avoid contamination with saliva. The individual rumen fluid samples were squeezed through 4 layers of cheesecloth; the pH was measured immediately, and then 6 mL each of strained fluid was acidified with 3 mL of 0.5 mol/L HCl and frozen at -20 ℃ for ammonia nitrogen (NH3-N) analysis . A 4-mL aliquot from each sample was prepared for volatile fatty acid (VFA) analysis using gas chromatography as described by Erwin et al..
Blood samples were taken from the external jugular vein of each calf 2 hours after the morning feeding of the liquid feed at 14, 28, and 60 days of age. At each collection, a duplicate 10 mL of blood samples were placed into tubes containing no additives. Serum was prepared by centrifugation at 3 000 ×g for 15 min at 4 ℃ and then stored at -20 ℃ until analysis. Activities of glutathione peroxidase (GSH-Px), superoxide dismutase (SOD), and concentration of maleic dialdehyde (MDA) were analyzed using respective commercial kits (Nanjing Jian Cheng Bioengineering Institute, Nanjing, China) as described previously . The inter-assay CVs of SOD, MDA, and GSH-Px were 1.57, 3.28, and 3.21, respectively, and the intra-assay CVs of SOD, MDA, and GSH-Px were 3.64, 4.29, and 4.51, respectively. The serum concentrations of immunoglobulin (Ig) A (IgA), IgG, and IgM were determined using ELISA kits (F4042-A, F3995-A, and F6685-A, respectively) (Shanghai Panke Industrial Co., Shanghai, China). The inter-assay CVs of IgG, IgA, and IgM were 7.30, 6.89, and 5.34, respectively, and the intra-assay CVs of IgG, IgA, and IgM were 8.69, 9.23, and 8.06, respectively.
Samples of MR and starter were collected once per week, and the content of DM, crude protein (CP), ash, ether extract (EE), acid detergent fiber (ADF), neutral detergent fiber (NDF), calcium (Ca), and phosphorus (P) were determined as described in a previous study . The content of protein, fat, lactose, total solids, and non-fat solids of the milk was determined with the mid-infrared procedures using a Milk Oscan Minor machine (MilkoScan Type 78110; Foss Electric, Hillerød, Denmark).
Statistical analysis was performed using SAS v. 9.4 (SAS Institute, Cary, NC) with all data tested first for normality. Linear, quadratic, and cubic polynomial contrasts were tested using the CONTRAST statement of SAS. The statistical model for ADG, liquid feed DM intake, starter DMI, total DMI, and F:G ratio included calf as random effect, treatment, day, and interaction between treatment and day as fixed effects, day as repeated effect, initial BW and parity of the dams as covariate. The statistical model for rumen fermentation characteristics, MDA, CAT, GSH-Px, SOD, IgA, IgM, and IgG included calf as random effect, treatment, day, and interaction between treatment and day as fixed effects, day as repeated effect, and parity of the dams as covariate. The statistical model for final BW, withers height, body length, and heart girth included treatment as fixed effects with initial BW, initial withers, initial body length, initial heart girth and parity of the dams as covariate. The covariance structure was autoregressive (1) based on Akaike's Information Criterion (AIC) and Schwarz's Bayesian Criterion (SBC).
Differences were considered as statistically significant when P < 0.05, and a tendency was considered when 0.05 ≤ P ≤ 0.10. Statistical power analysis was performed with α= 0.05 and power = 0.80 using PROC POWER procedure of SAS (SAS Institute Inc., Cary, NC), and with 10 calves per treatment group, a 10% difference between treatment means for most variables would be found with a power of 80% or greater.