Soybean oil supplementation and alfalfa hay inclusion in starter feed of Holstein dairy calves: growth performance, digestibility, ruminal fermentation and urinary purine derivatives

Abstract Forty newborn Holstein female calves (BW = 39.9 ± 2.1 kg) were assigned to 1 of 4 treatment groups (each consisting of 10 animals) in a 2 × 2 factorial arrangement of supplemental soybean oil [0 vs. 3% soybean oil (SBO) on dry matter basis] and forage level [0 vs. 15% alfalfa hay (AH) on dry matter basis] to evaluate the interaction effect of supplemental fat and forage feeding level in starter feed of dairy calves. Treatments were; (1) neither SBO supplementation nor AH inclusion (NSBO-NAH); (2) SBO supplementation but no AH included (SBO-NAH); (3) no SBO supplementation but AH included (NSBO-AH); and (4) SBO supplementation with AH (SBO-AH). Calves had ad-libitum access to water and starters throughout the study and weaned on day 63 of age but remained in the study until day 73 of age. The results showed that SBO supplementation reduced starter intake, average daily gain (tendency) and faecal consistency compared to un-supplemented diets. The lowest digestibility for neutral detergent fibre and crude protein, as well as the lowest wither height and volatile fatty acid production, were found for SBO-AH diet among experimental treatments. Moreover, the lowest urinary purine derivatives excretion but the highest urinary nitrogen excretion found in SBO-AH diet indicated the lowest nitrogen utilisation efficiency among experimental treatments. In summary, based on the current study condition, because the negative effects of SBO supplementation were exacerbated when AH was incorporated in the starter feed, concurrent feeding of SBO and AH is not recommendable in young calves. HIGHLIGHTS Supplemental soybean oil and alfalfa hay inclusion in starter feed of dairy calves was evaluated. Supplementation of SBO reduced crude protein digestibility and ruminal propionate concentration and caused looser faecal consistency. The inclusion of AH exacerbated the negative effects of SBO supplementation on ruminal fermentation and microbial activity. Concurrent supplementation of SBO and AH inclusion is not recommendable in dairy calves.


Introduction
Grains and fats are the main sources of energy in ruminants' diets. There are some limitations for using high-grain diets because of acidosis incidence and dairy calves had lower ruminal pH compared to mature ruminates (Laarman and Oba 2011) and fat supplementation can be a strategy to reduce ruminal acidosis risk due to reduced grain feeding (Stewart and Schingoethe 1984). Fat supplementation has been shown to be favourable for dairy calves in some studies (Ballou and DePeters 2008;Hill et al. 2011;Ghasemi et al. 2017); however, some others reported lower performance with fat supplementation (Kuehn et al. 1994;Hill et al. 2015;Ghorbani et al. 2020;Yousefinejad et al. 2021). Fat source and level (Miller et al. 1959;Ghorbani et al. 2020), environmental temperature (Ghasemi et al. 2017), the delivery method in starter feeds (Berends et al. 2018) and fatty acid (FA) profile (Quigley et al. 2019) are main factors that have an impact on responses. Fat supplementation may have an interactive effect with other nutrients in the diet. For instance, fat supplementation reduced fibre (Fallon et al. 1986) or protein digestibility (Hill et al. 2015). Although it has been stated that supplemental fat had a detrimental effect on fibre digestibility in mature ruminants which was mostly due to the coating effect of fat on fibre digestion (Soliva et al. 2004;Maia et al. 2010); however, the interaction of supplemental fat with dietary fibre content in dairy calves that have less ability to fibre digestion is still uncertain.
Fibre level in the starter is an important factor influencing dairy calves' well-being. Increased rumen motility, promote rumination, improve the integrity and healthiness of the rumen wall and reduce behavioural problems are the beneficial aspects of allocating the forage in starter feed in the pre-weaning period (Suarez et al. 2007;Beiranvand et al. 2014;). On the other side, displace concentrate intake and shift rumen fermentation in favour of acetate rather than butyrate, delay rumen papillae development, reduce starter feed intake and decrease body weight and dry matter digestibility are the unfavourable outcomes in forage-included starter feeds (Nocek and Kesler 1980;Phillips 2004). Moreover, forage inclusion reduces energy density per unit of starter feed due to lower energy content of forage than concentrate fraction (Beiranvand et al. 2014). Therefore, some negative effects observed in forage included starter feeds may arise in the shadow of the low energy content of starters. It could be postulated that supplemental fat, with a high energy level supplied in the diet, may compensate for the lower energy content in forage-included starter diets. However, supplemental fat, on the other side, may have some toxic effects on rumen microorganisms, adhere to the feed particles and create a physical barrier that makes an obstacle for microbial activity and fibre digestibility (Palmquist and Jenkins 1980;Maia et al. 2010). However, this hypothesis needs to be more evaluated in pre-ruminant animals. More research is warranted to be conducted regarding the concurrent feeding of fat along with forage in dairy calves.
We hypothesised that supplementing soybean oil as a fat source may compensate for the lesser energy content supplied through forage incorporation in the starter feeds. The objective of the current study was to evaluate the effects of different levels of SBO (0 vs. 3%, DM basis) and AH inclusion level (0 vs. 15%, DM basis) on the performance, structural growth, ruminal fermentation characteristics, nutrient apparent digestibility and urinary purine derivatives in young calves.

Calves and management
The present study was conducted on a commercial dairy farm (Avin-Dasht Dairy Industry Co, Qazvin, Iran). A total of forty 3-day-old Holstein female dairy calves with 39.9 (± 2.1) kg of initial BW were randomly assigned to a completely randomised design with a 2 Â 2 factorial arrangement of treatments (10 calves/ treatment) with the factors of soybean oil supplementation (0 vs. 3%, dry matter basis) and alfalfa hay inclusion levels (0 vs. 15%, dry matter basis). Calves were separated from their dams shortly after birth, housed in individual pens (1.3 Â 2.5 m) which were similar in design and bedded with sand that renewed every 24 h. The vaccination schedule and rearing system protocols were as conventional farm protocol. The calves fed 5 L of colostrum within the first 12 h of life (2.5 L of colostrum within 2 h of life and 2.5 L in a second feeding). The calves received 4 L of whole milk/day from day 3 to 10, 7 L/day from day 11 to 53 and 3 L/ day from day 54 to 63 in galvanised tin buckets (twice daily at 09:00 and 18:00 h). The average composition of offered milk was 3.16 ± 0.09% fat, 3.05 ± 0.06% crude protein (CP), 4.84 ± 0.06% lactose, and 11.9% total solids. Calves in all experimental treatments were weaned on day 63 of the study, but the experimental diets were continued 10 days later until day 73 of the study. Calves had free access to water during the whole experimental period.

Experimental treatments
Starter feeds were formulated to meet the National Research Council (2001) recommendations and were different in fat and fibre level. The experimental treatments were: (1) neither soybean oil supplementation nor alfalfa hay inclusion (NSBO-NAH), (2) soybean oil supplemented with no alfalfa hay inclusion (SBO-NAH), (3) no soybean oil supplementation but alfalfa hay included (NSBO-AH) and (4) soybean oil supplemented with alfalfa hay included (SBO-AH). The SBO source (Naz Industrial Vegetable Oil Co., Isfahan, Iran) was with the following fatty acids compositions: C16:0 ¼ 12.1%, C18:0 ¼ 5.2%, C18:1 ¼ 21.8%, C18:2 ¼ 51.2%, C18:3 ¼ 8.1% and other fatty acids ¼ 1.6%. Calves were fed basal starters in meal-form with an average geometric mean particle size of 0.74 ± 0.1 mm (American Society of Agricultural Engineers 1995). Alfalfa hay was chopped to obtain the geometric mean particle size equal to 2.8 ± 0.12 mm. After chopping, it was well-mixed with starter feed in forage-supplemented groups before feeding to dairy calves, and then the particle size distribution of starter feed was re-measured which was 0.89 ± 0.2 mm. The composition of the starter concentrates was kept constant within treatment, before and after weaning. The ingredients and chemical composition of experimental diets are presented in Table 1.
Starter intake, performance, faecal score and digestibility The amounts of starter diets offered (at 08:00 h) and refused were recorded (at 07:30 h) daily throughout the experiment. Measurement of BW was taken at 10day intervals during the experimental period using an electronic balance. Average daily gain (ADG) was calculated as the difference between BW taken every 10 days apart divided by 10. Feed efficiency (FE: kg of BW gain/kg of total dry matter intake) was also calculated in a 10-day interval. Total DMI was considered as liquid feed DMI þ starter feed DMI. Samples collected from feeds and orts were dried in a convection oven (60 C for 48 h). Subsamples of dried feeds and orts were well mixed and ground in a mill (Ogaw Seiki CO., Ltd., Tokyo, Japan) to pass a 1-mm screen and were analysed for CP (method 988.05; Association of Official Analytical Chemists 2002) ether extract (method 920.39; Association of Official Analytical Chemists 2002), and NDF without sodium sulphite, but with the inclusion of a-amylase (Van Soest et al. 1991). The non-fibrous carbohydrate (NFC) component was calculated based on National Research Council (2001) equation as follow; 100 À (CP þ NDF þ EE þ ash).
Faecal samples were collected via rectal palpation at 6 and 18 h after the morning meal (10 samples obtained for each animal) (Rastgoo et al. 2020). Faecal samples were dried in a forced dried oven (60 C; 48 h), and then ground in a Wiley mill through a 1mm screen. Aliquots of all faecal samples collected for each calf were mixed to obtain one composite sample for each animal. The composite faecal samples then were analysed to determine total nitrogen, EE, ash and NDF. Apparent total tract digestibility of nutrients (OM, NDF, CP and EE) was measured by using acid insoluble ash (AIA) as an internal marker. Apparent nutrients digestibility was calculated based on the concentrations of these nutrients and AIA in the feed (corrected for refusals) and faecal samples using the following formula: AD (%) ¼ 100 À 100 Â (MD/MF) Â (NF/ND), where AD is the apparent digestibility (%), MD is the marker in the diet (%), MF is the marker in the faeces (%), NF is the nutrient in the faeces (%) and ND is the nutrient in the diet (%) (Van Keulen and Young 1977).

Growth indices recording
The growth indices including heart girth, body length, body girth, withers height, hip height and hip-width were taken at day 3 (Initial), day 63 (weaning), day 73 (the final day of measurements) of age in the morning and before feeding based on the method described by Khan et al. (2007) for dairy calves. (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). b Contained per kilogram of supplement: 500,000 U of vitamin A, 100,000 U of vitamin D, 500 U of vitamin E, 7500 mg of Mn, 100 g of Ca, 8250 mg of Zn, 30 g of P, 20.5 g of Mg, 20 g of Na, 60 mg of Co, 2375 mg of Cu, 56 mg of I and 40 mg of Se. c Values were chemically analysed in laboratory. d Calculated from National Research Council (2001). e Starch content of experimental diets were calculated based on Cornell Net Carbohydrate and Protein System (Lanzas et al. 2007; CNCPS, v. 6.1).

Ruminal sampling and chemical analysis
Rumen fluid (30 mL) was collected on day 35 (preweaning) and day 70 (post-weaning) of the experiment using a stomach tube fitted to a vacuum pump 3-4 h after morning feeding; the first 10 mL was discarded because of possible saliva contamination and rumen pH was measured immediately (HI 8314 membrane pH metre; Hanna Instruments, Villafranca, Italy). The rumen samples were squeezed through 4 layers of cheesecloth. A 10-mL aliquot was preserved with 2 mL of 25% metaphosphoric acid and frozen at À20 C until analysis for volatile fatty acids (VFA). After thawing at room temperature, the rumen samples were analysed for VFA using gas chromatography (model CP-9002; Chrompack, Delft, the Netherlands) with a 50 m (0.32 mm ID) silica-fused column (CP-Wax Chrompack Capillary Column; Varian, Palo Alto, CA) as previously described .

Urine sampling and microbial protein synthesis measurements
The microbial protein synthesised in the rumen was estimated through purine derivatives (PD) obtained via spot sampling technique in post-weaning calves. As milk contains PD (Gonzalez-Ronquillo et al. 2003) and could cause an error in estimating PD excretion results obtained in pre-weaning calves, the spot urine sampling technique was used for MPS estimation only in the post-weaning period when calves received no milk as described by Makizadeh et al. (2020). Urine volumes were estimated as BW Â 26.8/creatinine concentration (mg/L) in post-weaned dairy calves as reported by Dennis et al. (2018). Spot urine samples were collected on four consecutive sampling days (from 69 to 72) during the post-weaning period from each animal. Samples were collected when calves urinated spontaneously ($10 mL). An aliquot of 5 mL of each sample was diluted immediately with 45 mL of 0.036 N sulphuric acid and stored at À20 C for analysis.
Later, urine samples were thawed at room temperature and analysed to determine the creatinine (Kit No. 555-A; Sigma Chemical Co.), and uric acid (Kit No. 685-50; Sigma Chemical Co) using spectrophotometer (UV-2600i, Shimadzu, Japan) and UN (using the assay described by Broderick and Kang 1980). Allantoin was measured using high-performance liquid chromatography by the method described by Chen and Gomes (1992). Total excretion of allantoin and uric acid was calculated from estimated daily urine output and determined metabolite concentrations.
The ruminal microbial N synthesis was calculated from daily urinary PD output using the following equation described by Chen and Gomes (1992): Microbial nitrogen (g N/day) ¼ X (mM/day) Â 70/ (0.116 Â 0.83 Â 1000); where X is microbial purine absorbed (mM/day), 70 is the N content of purines coefficient (mg N/mM), 0.116 is the ratio of purine-N to total N in mixed ruminal microbes which is 11.6: 100, and 0.83 is average digestibility of microbial purines (Chen and Gomes 1992).

Statistical analysis
Statistical analyses were conducted for 3 periods: preweaning (days 3-63), post-weaning (days 64-73) and the entire period (days 3-73) using PROC MIXED of SAS (version 9.1; SAS Inst. Inc., Cary, NC). The following model was adopted: where Y ijk is the dependent variable; m is the overall mean; SBO i is the effect of soybean oil supplementation levels (i ¼ 0 vs. 3%, DM basis); AH j is the effect of alfalfa hay inclusion (j ¼ no alfalfa hay supplementation and 15% alfalfa hay, DM basis); P k is the effect of period; (SBO Â P) ij is the interaction between soybean oil supplementation and period; (AH Â P) ik is the interaction between alfalfa hay inclusion and period; (SBO Â AH) jk is the interaction between soybean oil supplementation and AH inclusion; (SBO Â AH Â P) ijk is the tripartite effect of soybean oil supplementation, AH inclusion and period; b(X i À X ̅ ) is the covariate variable and e ijkm is the overall error term. The model contained calf within treatment as a random effect and the first-order autoregressive covariance structure (AR1) was determined as the most appropriate covariance structure for all repeated statements according to Akaike's information criterion and Bayesian information criterion. The BW and growth indices on the initial day of the experiment (day 3) were used as a covariate for weaning and final measurements of related items. Effects were considered to be significant when p .05 and it has been considered to have tendency was considered when .05 < p .10. All reported values through the tables are least-squares means.

Starter intake and performance
The results showed that the highest and the lowest starter intake during the pre-weaning period were observed in NSBO-AH and SBO-AH treatments, respectively (p < .05; Table 2). Total DMI tended to be lower in calves that received SBO-AH as compared with other groups (p ¼ .06). The ADG tended to be lower for SBO-AH in comparison with other treatments during the pre-weaning period (p ¼ .08). Accordingly, the lowest BW at weaning time was found for SBO-AH treatment when compared with others (p < .05). AH inclusion in starter feed reduced ADG and FE during pre-weaning and the entire period (p < .05).

Health indices and digestibility
The supplementation starters with SBO increased faecal score during the early 3 weeks as well as the entire period of the experiment compared with un-supplemented starters (p < .05; Table 3). The body temperatures of experimental calves were not influenced by experimental treatments in an experimental period (p > .05).
The interaction between SBO and forage level was significant in the current study, where the lowest digestibility of CP (p ¼ .04) and NDF (p ¼ .05) was found for SBO-AH. Digestibility of OM was reduced in claves fed AH (p ¼ .01) and tended to be reduced in calves supplemented with SBO (p ¼ .07).

Structural growth
Results showed that the calves fed SBO-AH diet had the lowest wither height, both in weaning time and final measurement (p < .05; Table 4). Body length, heart girth, body barrel and hip-width were not influenced by SBO supplementation, forage feeding level, or their interaction (p > .05). Hip height is reduced (p ¼ .05) when calves are supplemented with SBO or fed with AH (p < .01).

Ruminal fermentation profile
The lowest ruminal concentration of VFA (p < .05) during the pre-weaning period was found in calves that received SBO-AH treatment ( Table 5). The lowest acetate concentration during the post-weaning period was found for NSBO-NAH treatment (p < .05). The supplemental SBO reduced ruminal concentrations of propionate in pre-weaning (p ¼ .01) and valerate in post-weaning (p ¼ .05) periods. Inclusion of AH in starters increased ruminal acetate but reduced ruminal propionate and butyrate concentrations (p < .05).

Urinary purine derivatives and urinary nitrogen excretion
Results show that TPD, MPS and UN were influenced by the interaction of SBO supplementation and AH inclusion in starter with the SBO-AH treatment showed to have the lowest TPD and MPS (Table 6). However, the greatest UN excretion among experimental treatments was found for SBO-AH treatment (p < .05). Table 2. Least square means for starter intake, average daily gain and feed efficiency in dairy calves supplemented with soybean oil (0 vs. 3%, DM basis) with different alfalfa hay levels (0 vs. 15%, DM basis) in starter feed (n ¼ 10 calves per treatment). (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). 2 Statistical comparisons -SBO: soybean oil supplementation level in the starter (0 vs. 3%); AH: alfalfa hay inclusion level in the starter (0 vs. 15%); SBO Â AH: interaction between soybean oil supplementation and alfalfa hay inclusion levels in the starter. 3 Kilogram of body weight gain/kg of total dry matter intake. Means within a row with different superscript letters are different (p < .05).
Urinary allantoin concentration was also tended to be reduced in SBO-AH treatment (p ¼ .08).

Interaction of SBO supplementation and AH inclusion
The lowest starter intake during pre-weaning period was observed in SBO-AH treatment (494 g/day) compared to other treatments. It has been suggested that the lower feed intake in fat supplemented diets may be related to the decreased palatability and reduced nutrient digestibility (Ghorbani et al. 2020). Furthermore, in dairy cows, a negative effect of fat called 'greasiness' on intake has been proposed by Drackley et al. (1994). In addition, fatty acid oxidation in the liver is indicated to have the potential to control the appetite in mature ruminants (Harvatine and Allen 2005). Some studies indicated the inflammatory effect of some individual fatty acids in dairy calves that can be related to reduced starter intake (Tsai et al. 2017). In addition to the unfavourable effect of SBO on starter intake, intake was more exacerbated when AH was included in diet accompanied with SBO supplementation. This is probably related to under-developing ruminal conditions during the pre-weaning period (Cersosimo et al. 2019;Yousefinejad et al. 2021). Pre-weaning dairy calves having a limited cellulolytic activity that could further negatively influence ruminal microbes' activity and consequently fibre digestibility when an unsaturated fatty acid source such as linseed oil (Ikwuegbu and Sutton 1982) or soybean oil (Ghorbani et al. 2020) was supplemented in the diet. Physical coating of the fibre by fat or toxic effects of fat on ruminal microbes may be responsible in part for the lower digestibilities of fibre and protein found in the SBO-AH group (Soliva et al. 2004;Maia et al. 2010). In agreement with previous works (Hill et al. 2015;Ghorbani et al. 2020;Yousefinejad et al. 2021) our results suggest that SBO supplementation was associated with reduced digestibility of OM and NDF in dairy calves, which was more negatively influenced when forage was included in the starter. Regarding the interaction between SBO supplementation and AH inclusion in the starter feed in SBO-AH treatment, it can be postulated that incorporating the AH increased NDF content and SBO supplementation reduced NDF digestibility, from the other side indicating that supplemental SBO can be more detrimental for NDF digestibility when forage is included in starter feed and the least NDF digestibility is found to be SBO-AH treatment. Reduced digestibility of nutrient can be a factor to reduce starter intake in dairy calves (Ghorbani et al. 2020).
In the current study, the lowest ADG (554 g/day) during pre-weaning period was found in SBO-AH Table 3. Least square means for faecal score, rectal temperature and nutrients digestibility in dairy calves supplemented with soybean oil (0 vs. 3%, DM basis) with different alfalfa hay levels (0 vs. 15%, DM basis) in starter feed (n ¼ 10 calves per treatment). (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). 2 Statistical comparisons -SBO: soybean oil supplementation level in the starter (0 vs. 3%); AH: alfalfa hay inclusion level in the starter (0 vs. 15%); SBO Â AH: interaction between soybean oil supplementation and alfalfa hay inclusion levels in the starter. Means within a row with different superscript letters are different (p < .05).
calves, which coincided with a similar trend in BW and wither height. These changes along with the lower digestibility of NDF and CP in SBO-AH calves might reflect a less efficient tissue accretion in SBO-AH calves compared to other treatments.
In addition to the lower nutrient digestibility, the lower concentrations of VFA seem to be responsible for the lower growth performance observed in calves fed SBO-AH diet. The lower starter intake and the lower NDF and CP digestibilities in SBO-AH group may reduce substrate for ruminal microbial fermentation, and hence VFA concentration was reduced in the rumen. In ruminants, considerable energy requirement (approximately 70%) is supplied through VFA produced in the rumen (Bergman 1990); thus, the lower ruminal VFA concentration in the calves fed SBO-AH would be expectedly accompanied by a lower supply of energy for growth. The least acetate concentration (43.9 mM) but the highest propionate concentration (33.4 mM) were observed in NSBO-NAH treatment in the current study indicating the greater energy supplied in calves fed this diet compared to other experimental diets. Previous works stated that increased forage level in the starter feed can positively influence acetate concentration . The lowest acetate concentration is found when starter feed was neither contained AH nor SBO indicating that the inclusion of AH and supplemental SBO both shift the rumen fermentation towards more acetate rather than propionate in dairy calves. As discussed earlier, more propionate is favourable than more acetate for rumen development (Bergman 1990); hence, supplemental SBO along with AH inclusion in starter feed is not recommendable from the ruminal fermentation perspective.
In the current study, supplementation of SBO along with forage inclusion reduced total PD excretion, and thus reduced estimated MPS, but increased UN excretion. This indicates lower nitrogen utilisation efficiency in SBO-AH treatments compared to other groups. It has been indicated before that the amount of total PD excreted is influenced by feed intake (Singh et al. 2007). Therefore, the lower starter intake in SBO-AH group had a substantial role in reducing total PD excretion. In addition, lower PD in the SBO-AH diet was partially due to decreased digested NDF and CP in this treatment. It has been indicated that PD excretion can be an indicator of ruminal development in Table 4. Least square means for growth indices in dairy calves supplemented with soybean oil (0 vs. 3%, DM basis) with different alfalfa hay levels (0 vs. 15%, DM basis) in starter feed (n ¼ 10 calves per treatment). (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). 2 Statistical comparisons -SBO: soybean oil supplementation level in the starter (0 vs. 3%); AH: alfalfa hay inclusion level in the starter (0 vs. 15%); SBO Â AH: interaction between soybean oil supplementation and alfalfa hay inclusion levels in the starter. Means within a row with different superscript letters are different (p < .05). dairy calves (Terr e et al. 2006). Therefore, although with no ruminal morphological measurements in the current study, our results indicate that dairy calves fed the SBO-AH diet obtained less developed rumen in comparison with other diets. With respect to the interaction of SBO supplementation and AH inclusion in starter feed, results indicate that microbial activity was the highest when neither SBO was supplemented nor AH was included ins starter feed indicating that highfat diet along with high fibre content in starter diet cannot provide adequate energy required for microbial development. Higher UN excretion in the SBO-AH diet compared with other diets is probably related to lower utilisation of ruminal ammonia-N. Fat supplementation has a direct effect on the reduction the ruminal microbial activity (Nagaraja et al. 1997;Fiorentini et al. 2013). To the best of our knowledge, the current study is the first report regarding the interaction effect of SBO supplementation and starter forage level on urinary PD excretion and urinary nitrogen excretion in dairy calves. More research with different sources of fat and fibre contents in starter feeds may give more insight into the underlying mechanisms, which can be used to optimise fat and forage feeding strategies in dairy calves that will positively impact calf growth and future outcomes (Van De Stroet et al. 2016).

Effect of SBO supplementation
Reduction in starter feed intake was reported by Hill et al. (2015) when 2% SBO was included in diets for dairy calves (before 8 weeks of age). The effect of fat supplementation on starter intake in dairy calves might be related to the type of fat supplements (e.g. whole or extruded oilseeds), degree of saturation of dietary fat, supplemented level in the starter feed (Hill et al. 2011(Hill et al. , 2015Ghasemi et al. 2017), the fat delivery method in starter feed (Berends et al. 2018), and supplemented FA profile (Quigley et al. 2019). Our result adds to our knowledge that in addition to the items mentioned, forage level in diet also can be an important item influencing the animal response to supplemental fat. Starter intake in SBO-NAH group is 612 g/ Table 5. Least square means for ruminal metabolism in dairy calves supplemented with soybean oil (0 vs. 3%, DM basis) with different alfalfa hay levels (0 vs. 15%, DM basis) in starter feed (n ¼ 10 calves per treatment). (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). 2 Statistical comparisons -SBO: soybean oil supplementation level in the starter (0 vs. 3%); AH: alfalfa hay inclusion level in the starter (0 vs. 15%); SBO Â AH: interaction between soybean oil supplementation and alfalfa hay inclusion levels in the starter. Means within a row with different superscript letters are different (p < .05). 3 Branched-chain volatile fatty acids (BCVFA) are the molar proportion of valerate þ isovalretae. day and reduced to 494 g/day when SBO was supplemented along with AH inclusion in starter feed indicating that concurrent feeding of SBO and AH in young calves may not be favourable.
Supplementation of starter feeds with SBO reduced faecal consistency during early weeks of life. This could indicate that SBO supplementation could increase the passage rate in the gastrointestinal tract, thus causing looser faeces (Ghorbani et al. 2020). No difference in feeding different fat sources on the faecal score in dairy calves was observed in a cold environment (Ghasemi et al. 2017). The CP digestibility was influenced negatively by SBO supplementation in the current study. Hill et al. (2015) reported that calves under 2 months of age fed a starter feed supplemented with SBO had lower digestibility of OM and CP, but not in tallow supplemented calves. Interestingly and contrary to our results, Araujo et al. (2014) reported that CP digestibility was increased with supplemented fat. These inconsistent results may be related to the fat level and source (saturated vs. unsaturated) as well as the ruminal accessibility of the supplemented fatty acids. For instance, it has been stated that unsaturated FA source (linseed oil in Ikwuegbu and Sutton 1982; soybean oil in Ghorbani et al. 2020) has a detrimental effect on nutrient digestibility due to the ruminal accessibility of FA in the rumen; however, inaccessibility of FA source in the rumen (calcium salt of linseed oil in Kazemi-Bonchenari et al. 2020) had been shown to have no effect on nutrient digestibility.
Ruminal acetate concentration was increased but ruminal propionate concentration was reduced when SBO was supplemented in calves. Propionate supplies greater energy on a molar basis than acetate (Bergman 1990), and thus NSBO diet has greater potential compared to the SBO diet to improve animal performance. This may eventually cause lower hip height in calves at weaning time when supplemented with SBO. Lower CP digestibility in SBO supplemented diets may be contributed to reducing BCVFA concentration due to the lower branched amino acids available that are precursors for BCVFA in ruminal fluid (Yang 2002).
The supplemented SBO reduced purine derivatives excreted through urine indicating lower microbial activity in fat-supplemented calves. Supplemental SBO increased the UN indicating lower nitrogen efficiency in these diets (Kauffman and St-Pierre 2001). The effect of supplemental fat which is ruminally available on ruminal microbes (Nagaraja et al. 1997;Fiorentini et al. 2013) could alter nitrogen utilisation efficiency. In addition, lower OM digestibility in the SBO supplemented calves resulted in less nitrogen incorporation in ruminal protein biosynthesis and then less efficient nitrogen utilisation. The present study results confirm that from the nitrogen metabolism viewpoint, feeding low-fat starter feeds are more efficient than fat supplemented starter feeds in young calves.

Effect of AH inclusion
Forage inclusion level in the starter feed reduced BW as well as FE at final measurement. Moreover, consistent with BW results, hip height was reduced in claves fed forage in starter feed. This is partly related to reduced OM digestibility as well as reduced VFA concentration in AH included diets compared to other treatments. As stated by Bergman (1990), lower VFA concentration as the main energy source in ruminants can result in less energy towards animal growth. Furthermore, lower ruminal propionate and butyrate concentrations in forage included diets along with greater ruminal acetate concentration clarify the lower Table 6. Least square means for purine derivative excretion and microbial protein synthesis in dairy calves supplemented with soybean oil (0 vs. 3%, DM basis) with different alfalfa hay levels (0 vs. 15%, DM basis) in starter feed (n ¼ 10 calves per treatment). (1) no soybean oil supplementation with no forage included in the starter (NSBO-NAH); (2) no soybean oil supplementation with 15% alfalfa hay included in the starter (NSBO-AH); (3) 3% soybean oil supplemented with no forage included in the starter (SBO-NAH); (4) 3% soybean oil supplemented with 15% alfalfa hay included in the starter (SBO-AH). 2 Statistical comparisons -SBO: soybean oil supplementation level in the starter (0 vs. 3%); AH: alfalfa hay inclusion level in the starter (0 vs. 15%); SBO Â AH: interaction between soybean oil supplementation and alfalfa hay inclusion levels in the starter. Means within a row with different superscript letters are different (p < .05). energy supplied in forage-fed calves. It has been identified that propionate and butyrate are the two main VFA that supply greater energy on a molar basis than acetate (Bergman 1990), and thus provide greater energy to improve animal performance. Some of the beneficial effects of forage inclusion in starter feed have been identified in pre-ruminant animals such as stimulating effect of forage for the muscular layer of the rumen and promote rumination (Phillips 2004), maintain the integrity and healthiness of the rumen wall, and increase rumen motility and prevent hyperkeratinization of ruminal papillae (Suarez et al. 2007). Some discourage viewpoints also were indicated that forage may displace concentrate intake and shift rumen fermentation in favour of acetate rather than butyrate production and, thus, delay rumen papillae development (Tamate et al. 1962). The amount of forage included in starter feed is supposed to have a pivotal role in pre-weaned calves' responses to forage feeding (Suarez et al. 2007;Mirzaei et al. 2016). Some previous studies stated that the minimum amount of forage in starter feed to keep to avoid ruminal pH reduction is 5% (DM basis) (Aragona et al. 2020). We found that in addition to forage feeding level and other related variables, supplemental fat also can impact the responses observed in forage-fed calves. For instance, forage inclusion per se did not have a negative impact on starter intake and NDF digestibility; however, these items were negatively influenced when forage was fed along with SBO. Thus, it can be stated here that the supplemented fat level and the source may be considered while identifying the optimum forage inclusion in dairy calves' starter feeds.

Conclusions
Soybean oil supplementation (3%, DM basis) concurrently with alfalfa hay inclusion (15%, DM basis) in the starter feed negatively influenced growth performance and digestibility of nutrients in young calves. Moreover, the nitrogen utilisation efficiency was negatively influenced by reducing the urinary PD excretion and increasing the urinary nitrogen excretion. In summary, considering variables evaluated in the current study, results indicated that soybean oil supplementation is not recommendable when alfalfa hay is incorporated in the starter feed of dairy calves.

Ethical approval
All experimental procedures were conducted according to the international procedures and approved by the University of Urmia Animal Ethics Committee.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
Thisstudy is only supported by Urmia University (PhD grant No. 94/269-1004) which pay forall PhD students and no fund was paid from the out of university for this study.

Data availability statement
Data available on request due to privacy/ethical restrictions.