The Effect of Different Feeding Times of Microencapsulated Sodium Butyrate in Whole Milk and Starter Feed on Growth and Health of Holstein Dairy Calves

This study aimed to determine the optimal feeding time of microencapsulated sodium butyrate (SB) in whole milk (WM) and starter feed on growth performance and health in dairy calves. Forty-eight Holstein calves (age = 4 d; body weight [BW] = 39.45 ± 2.48 kg) were randomly allocated to 1 of 4 treatment groups (12 calves per treatment) in a completely randomized block design and fed (1) WM without microencapsulated SB (SB0) supplementation; (2) 4 g/d SB added to WM since d 4 to 32 (SB4); (3) 4 g/d SB added to WM since d 61 to 74 and equal amount was added to starter since d 75 to 88 (SB60); and (4) 4 g/d SB added to WM (since d 4 to 74) and the starter (since d 74 to 88) throughout the experiment (SBT). Calves fed SB4 and SBT had lower fecal score during pre-weaning, transition period, and overall period (P = 0.043, P = 0.05, and P = 0.015; respectively). In addition, calves in SB4 and SBT groups decreased the number of days with scours during pre-weaning period, and throughout study (P = 0.035 and P = 0.025; respectively). SB60 calves had greater serum total protein (P < 0.001) during post-weaning period. Post-weaning and overall albumin concentrations were greater in SB4 and SBT calves (P = 0.011), and tended to increase in pre-weaning period compared to control calves (P = 0.06). Based on these results, addition of SB in WM is recommendable for the rst month of milk-fed calves life.


Introduction
Butyric acid, one of the short-chain fatty acids (SCFA), is a natural substance present in the rumen of ruminants, colons of monogastric species, cow milk (0.16 g/L); Alais, 1984;Guilloteau et al., 2010a). It has been well documented that butyrate is an important stimulator and regulator of the ruminal epithelium growth and function (Penner et al., 2011). In ruminants, the most important source of butyrate is microbial fermentation of carbohydrates in the rumen (Bergman, 1990). Within the rst 1 to 2 wk of a calf's life, solid feed intake is very low and rumen micro ora is not fully functioning. This leads to a very low butyrate concentration in the yet underdeveloped rumen until regular solid feed intake starts and rumen micro ora develops (Anderson et al., 1987;Flaga et al., 2015). Thus, prior to the development of rumen, milk butyrate is the main source of this molecule for the newborn calf. On the other hand, calves are fed mostly with whole milk (WM) or milk replacer (MR) before weaning and the abomasum and small intestine are the main sites of feed digestion; thus, the development of these gastrointestinal tract (GIT) compartments is crucial for nutrient absorption, performance and health of milk-fed calves. Because GIT development affects feed intake, e ciency of digestion, and resistance to gastrointestinal disorders, and thus animal growth and health, each method enhancing these processes is highly desirable. Therefore, supplementing liquid feed or starter feed with butyrate may be a good strategy to improve rumen and intestinal development in calves (Górka et al., 2011a).
Because they are more stable, generally odourless and easier to handle in the feed manufacturing processes, butyrate salts (sodium butyrate [SB] or calcium butyrate) or butyrins (esters of butyrate and glycerol) are often used instead of butyric acid itself in animal studies and in practice. (Guilloteau et al., 2010a). The SB, the most often used source of dietary butyrate due to its high availability and modest price, dissolves easily in water and rapidly dissociates in water solutions (Mallo et al., 2012). The effect of dietary butyrate supplementation can be modulated by butyrate protection from its release and utilization in the stomach (both forestomach and abomasum). Embedding in the continuous lipid matrix, often referred as microencapsulation or fat coating, is commonly used for this purpose ( advantage, protected butyrate is released slowly from the fat coat, providing the possibility for its more uniform distribution in the small and large intestine. Therefore, protected butyrate is more likely to affect the structure and function of the large intestine due to butyrate delivery to the very last sections of the intestine. This seems to be especially bene cial taking into account the high susceptibility of newborn calves to diarrhea (Gorka et al., 2018). It has been well demonstrated in numerous studies that butyrate supplementation in milk replacer and starter feed (Gorka et al., 2014(Gorka et al., , 2011a(Gorka et al., , 2011b(Gorka et al., , 2009Guilloteau et al., 2010bGuilloteau et al., , 2009bHill et al., 2007;Koch et al., 2019;Pouillart, 1998;Roh et al., 2018), and acidi ed milk (Sun et al., 2019) have pronounced effects on growth performance, feed e ciency, GIT development and health of dairy calves through modulation of proliferation, differentiation, stimulated pancreatic secretions, and function of the GIT tissues.
Despite the fact that butyrate is naturally found in cow's milk, it seems that because of its little amount, adding extra SB to milk can improve dairy calves' performance. To our knowledge, there is only one study in neonatal calves in which SB has been added to whole milk (Mahjoubi et al., 2020); where it has been indicated that SB (4 or 8 g/d) can improve calf performance but they also stated that the time of SB supplementation might have some effect on calf performance. Because of that, we decided to investigate this effect.
Based on above considerations, we hypothesized that the addition of SB to whole milk could improve calf growth performance and health status when it is added in rst month of life and or in transition period. Therefore, the main aim of the current study was to determine the optimum age for SB inclusion in whole milk.

Materials And Methods
This experiment was performed since January to April, 2021 in a commercial dairy farm (Avin Dasht, Qazvin, Iran). This farm is located in a sub-tropical area (longitude 49°29' E and latitude 35°57' N).

Calves, Treatments and Management
A total of 48 Holstein dairy calves (28 female and 20 male, average BW = 39.48 ± 2.48 kg) were randomly assigned to treatments (n = 12 calves per treatment; 7 female and 5 male) in a completely randomized block design (sex of animals was considered as a block). After birth, all calves were separated from their dams immediately and placed in individual pens (1.5 × 2.0 m 2 ) bedded with clean wheat straw, that was renewed every 24 h. Calves consumed at least 2.5 L colostrum within the 1 h after birth and 1.5 L fed 12 h after the rst feeding. The quality of colostrum was measured with a digital Brix refractometer (PAL-1, Atago Co. Ltd., Bellevue, WA; Brix% of 23-24). On d 2 and 3 of life, calves received transition milk (4 L) in two meals of equal volume (at 0800 and h). From d 4 onwards, calves were randomly assigned to experimental treatments and individually fed whole milk (3.51 ± 0.11% fat, 3.15 ± 0.1% crude protein, 4.67 ± 0.06% lactose, and 12.12 ± 0.17% total solids, 3.4 ± 0.7 g/100 g fatty acids butyric acid) two times daily at 0800 and 2000 h. Treatments were as follow: (1) control without microencapsulated sodium butyrate (SB0) supplementation; (2) 4 g/d SB (Novyrate ® C) added to milk since d 4 to 32 (SB4); (3) 4 g/d SB added to milk since d 61 to 74 and equal amount was added to starter since d 75 to 88 (SB60); and (4) 4 g/d SB added to milk (since d 4 to 74) and the starter (since d 74 to 88) throughout the experiment (SBT). Pre-weaning the SB was added to whole milk and mixed by a strew before calf drink and post-weaning was top-dressed to starter feed to make sure each calf eats the SB. Novyrate® C is a coated butyrate product (containing 32% sodium butyrate as microencapsulated and approximately 25% butyric acid, 99% dry matter, 61% crude fat, 20% ash, 6.4% sodium; Novyrate ® C, Innovad co, Essen, Belgium). All calves were fed the same volumes of whole milk in the galvanized tin buckets at the rate of 5 L/d from d 4 to 16 d of age; 7 L/d from d 17 to 59; 6 L/d from d 60 to 63; 5 L/d from d 64 to 66; 4 L/d from d 67 to 69 and 2 L/d from d 70 to 74 (only at the morning feeding) (Fig. 1). All calves were weaned at d 74 of age and stayed in individual stalls until d 88 of age to collect the postweaning data. The calves had free access to fresh starter feed and water every day. The starter feed with ground physical form was offered every morning at 0830 h throughout the study. Diet was formulated using the NRC (2001) software and nutrients composition of the starter feed are presented in Table 1. The calves received the starter feed mixed with 100 g/kg DM chopped alfalfa as a total mixed ration (TMR). The starter feed formulation was constant across experimental treatment during pre-and post-weaning periods.

Measurement of production performance and health
Throughout the study, feed offered and refused was weighed daily to determine the total starter intake for the individual calf. The day of age that each calf met speci c starter feed consumption target of 1 kg was recorded and used to evaluate the time to consume 1 kg starter feed. Individual body weight (using an electronic scale) and body skeletal growth including body length (distance between the point of shoulder and rump), withers height (distance from the base of the front feet to the withers), heart girth (circumference of the chest), and hip height (distance from the base of the rear feet to hook bones of the calves were recorded on d 4, 32, 60, 74, and 88 before the morning feeding meal, and ADG was calculated as the difference between two consecutive BW measurements divided by days. Gain-to-feed ratio was calculated as grams of ADG divided by grams of total DMI (liquid feed DMI + starter feed DMI). Samples of starter feed were collected throughout the study (n = 4, 1 sample per 20 d) for the determination of DM and chemical analyses. Samples of starter feeds were dried in a convection oven (60 °C for 48 h). Subsamples of dried feeds were composited by treatment and ground in a mill (Ogaw Seiki CO., Ltd., Tokyo, Japan) to pass a 1-mm screen. Health status was assessed daily according to Larson et al. (1977) and Heinrichs et al. (2003). One of the authors performed the health scoring and each time was the same person. Fecal scoring was as follows: 1 = rm, 2 = soft, 3 = soft and running, 4 = watery. General appearance scoring was: 1 = normal and alert; 2 = ears drooped; 3 = head and ears drooped, dull eyes, slightly lethargic; 4 = head and ears drooped, dull eyes, lethargic; and 5 = severely lethargic. For calves which need medical treatments, farm's veterinarian administrated the proper drug and the treatment was followed according to his recommendation; therefore, medical days, treatment bouts and number of used drugs were recorded to be statistically analyzed. Antibiotic therapy was initiated when rectal temperature was over 39.5°C.

Blood sampling and Analyses
Blood samples from each calf were collected from the jugular vein into 10 mL tubes 4 h after morning feeding on d 4, 32, 60, 74 and 88 (post-weaning). Blood samples were placed on ice immediately after collection and centrifuged at 3000 × g (KUBOTA Co., Bunkyo City, Tokyo, Japan) for 15 min at 4 °C to obtain serum, and then serum samples were frozen at -20 °C until future analyses. It took 1 hour from blood sampling to storage. Serum subsamples were analyzed to determine concentrations of glucose (mmol/l), albumin (g/dl) and total protein (TP, g/dl) using commercial kits (Pars Azmoon Co., Tehran, Iran). Serum concentrations of beta-hydroxybutyrate (BHB, mmol/l) were measured with a commercial kit (kit Ranbut, Randox Laboratories Limited, Crumlin, County Antrim, Randox, UK); the inter-and intra-assay coe cient of variation (CV) for the glucose assay were 2.34 and 2.72 %, respectively, and for the BHB assay were 2.91 and 3.45 %, respectively.

Statistical Analysis
Prior to data analysis, all data were evaluated for normality using the UNIVARIATE procedure of SAS (version 9.4; SAS Institute Inc., Cary, NC). Total DMI, starter feed intake, and blood metabolites data were subjected to ANOVA using the MIXED procedure of SAS with day as repeated measures during the overall experiment. The model consisted of treatment, sex, day, and their interactions were included as the xed effects and calf within treatment was included as a random effect. Sex was not signi cant and removed from the model. Initial BW, initial skeletal growth parameters, and blood data were considered as a covariate for the BW, skeletal growth, and blood metabolites analysis but were removed from the nal models because no differences due to these factors were found (P > 0.20). Body weight, gain-to-feed ratio, body skeletal growth data and health indices were not analyzed as repeated measure and they were analyzed by the GLM. Three variance-covariance structures (autoregressive order 1, unstructured or compound symmetry) were tested and autoregressive order 1 covariance structure yielded the smallest Schwarz's Bayesian information criterion. In addition to overall F test, differences among treatments were assessed using orthogonal contrast (SB0 vs. SB4, SB60 and SBT). The statistical model used for analysis was: where Y ijk = dependent variable; µ = overall mean; T i = xed effect of treatment i, D j = repeated measure of day j, (T × D) ij = xed effect of interaction between treatment and day, calf k = random effect of calf k ; β (Xi -X) is the covariate variable and e ijk = overall error term.
The number of days with diarrhea were categorized with fecal score ≥ 2 (Mahjoubi et al., 2017). General appearance score (1 to 5) was categorized in the number of days with general appearance score ≥ 2. The variance of number of days with fecal and general appearance score ≥ 2 were not uniformly distributed. Therefore, these variables were square-root transformed for better homogeneity of the distribution of residuals (means shown in Table 5 for these variables are back-transformed). The same was done for medical days, treatment bouts and number of used drugs. Least squares means for treatment effects were separated by the use of the PDIFF statement. Signi cance was declared at P ≤ 0.05 and tendencies at P ≤ 0.10.

Feed intake and growth performance
The results of DMI, starter intake, BW, ADG, and gain-to-feed ratio, and are shown in Table 2. In general, daily DMI and starter intake did not differ among treatment groups at any stage (P > 0.05). Supplementation with SB had no effect on BW, ADG (Fig.2), and gain-to-feed ratio (P > 0.05).
The results of the structural growth indices are given in Table 3. There was no effect of treatments on structural growth indices in different days of trial (P > 0.05).

Health criteria
Fecal scores and general appearance scores are given in Table 4. Compared to SB0 and SB60, encapsulated SB inclusion in SB4 and SBT groups decreased fecal score during pre-weaning (P = 0.043), transition period (d 61 to 88; P = 0.05) and throughout the experiment (P = 0.015). However, general appearance score did not differ among treatments at any stage (P > 0.05).
The number of days with loose fecal score (≥ 2) were lower (P = 0.035 and P =0.025; respectively) for calves SB4 and SBT compared to SB0 and SB60 groups during pre-weaning period and throughout the experiment, while did not differ among treatments during the rst month of life and transition period ( Table  4). In general, supplementation the encapsulated SB in WM signi cantly decreased (P = 0.039) the number of days with loose fecal score compared with control group during pre-weaning period. Throughout the study, calves fed SB tended (P = 0.07) to decrease the number of days with loose fecal score compared with control group. Days with altered general appearance score, medical days, treatment bouts and number of used drugs did not differ among treatments (P > 0.05; Table 5).

Blood metabolites
The results on the blood metabolites are presented in Table 6. Serum glucose levels were not different among treatments at any stage (P > 0.05). Serum BHB levels were not different among treatments during pre-and post-weaning period (P > 0.05), but there was a signi cant effect of interaction of treatment by day (P = 0.05) throughout the study. Serum BHB increased with the advancement in the study, but there was a tendency for SB4 calves to have higher (P = 0.09) BHB level than SB0 and SB60 calves at d 74. There was also difference at d 88 for SB60 and SBT calves to have higher (P < 0.001) BHB level than SB0 and SB4 calves. Throughout the study, level of serum albumin was higher (P = 0.011) in SB4 and SBT groups than the other groups. Calves fed encapsulated SB had higher serum albumin concentrations during pre-weaning, post-weaning and overall periods (P = 0.032 and P < 0.001; respectively) compared to calves without SB. Post-weaning TP was higher (P < 0.001) for SB4 and SBT groups than for SB0 and SB60 groups. Serum TP level in post-weaning period also was signi cantly higher (P = 0.025) for encapsulated SB fed calves than for SB0 calves.

Discussion
Feed intake and growth performance Most research has been done on calves in the pre-weaning period by adding butyric acid to formula or solid feeds. The effect of adding butyric acid to milk replacer on starter consumption has been contradictory  Health criteria Contrary to the initial hypothesis of this study, butyrate had no effect on diarrhea or loose feces during the rst month of life but is in agreement with our previous study (Mahjoubi et al.  (Górka et al., 2011a, 2011b, 2009). Less intestinal development, as a result of MR feeding instead of WM, makes newborn calves more prone to diarrhea (Blattler et al., 2001). Since WM was used in this study, it seems that the amount of butyric acid in milk has partially led to the development of intestinal tissue thereby to be able to handle diarrhea. It has been shown that when butyric acid is added to milk replacer, it improves colon function and can improve animal health (Guilloteau et al., 2009b). In a study to increase butyric acid in the rumen using molasses, the authors found that despite the increase in the concentration of butyric acid in the rumen, there was no effect on fecal score (Oltramari et al., 2016). Consistent with the present study, Wanat et al. (2015) also observed a linear effect with increasing butyric acid in the form of protected microcapsules on decreasing the fecal score. These results generally show that the effect of butyric acid addition to the starter depending to the level of supplementation and method of delivery which leads to contradictory results. In contrast with our expectation, although addition of SB in the WM in the rst month of life did not improve diarrhea during this time, it seems that observed improvement in fecal score is due to the long-term and carry-over effects of SB during pre-weaning and throughout study.

Blood parameters
Glucose is considered as the preferred energy substrate in pre-ruminant calves (Donkin and Armentano, 1995).
In agreement with previous studies (

Conclusion
Supplementation WM with SB did not have considerable effect on feed intake and growth performance of dairy calves during the pre-and post-weanig period. However, fecal score and number of days with loose feces decreased as SB was added to milk for early weeks of life (SB4) and through the entire period of study (SBT) during the pre-weaning, transition period, and throughout study. In conclusion, because there was no difference between SB4 and SBT, addition of SB to milk during rst month of life economically recommended.

Declarations Acknowledgments
The authors gratefully acknowledge Mr. Ashra nia and Dr. Mirzaei (Arak University, Arak, Iran), and Dr. Ghofrani (Javaneh Khorasan Company, Mashhad, Iran), and the staff of Avin Dasht Dairy Farm for their assistance in carrying out of this experiment. We also appreciate Javaneh Khorasan Company (Iran) for providing us with 25 kg of Novyrate ® C.

Funding
This work was developed as a part of the rst author's thesis (Grant No. 98-6994) that was nancially supported by deputy of research and technology at Arak University.

Con ict of interest
The authors declare no con icts of interest.

Ethics approval
All experimental procedures conducted in this study were approved by the Animal Care and Use Committee of Arak University (IACUC #IR2018011) following the guidelines outlined by Iranian Council of Animal Care (1995; #19356      Means within a row with different superscripts differ (P < 0.05). Means within a row with different superscripts differ (P < 0.05).