Introduction
Since 1970, ionophores such as sodium monensin (MON) have been used in feedlot cattle nutrition (Duffield et al., Reference Duffield, Merrill and Bagg2012). Such additives are used to prevent metabolic disorders and improve performance (Ellis et al., Reference Ellis, Dijkstra, Bannink, Kebreab, Hook, Archibeque and France2012). However, some European countries have banned the use of ionophores such as MON, classifying them as antibiotics (OJEU, 2003).
The European market was the second largest Brazilian beef importer with 270 419 tons in 2017 (ABIEC, 2016), and ionophores are the primary feed additives used in feedlot diets in Brazil (Oliveira and Millen, Reference Oliveira and Millen2014). This has stimulated research for alternatives to ionophores, which is the case of functional oils (FO). Benchaar et al. (Reference Benchaar, Calsamiglia, Chaves, Fraser, Colombatto, Mcallister and Beauchemin2008) reported that FO can be used to manipulate ruminal fermentation, improve nutrient utilization in ruminants and it is safe for human and animal health. Thus, the objective of this study was to evaluate the effects of adding FO or MON to feedlot diets on performance, carcass traits, feeding behaviour and rumen morphometrics of Nellore cattle.
Materials and methods
Treatments, feeding and management description
All the procedures involving the use of animals in this study were in accordance with the guidelines established by the São Paulo State University Ethical Committee for Animal Research (protocol number 22/2015).
The trial was conducted at São Paulo State University feedlot, Dracena campus, Brazil. Ninety-six 20-mo-old Nellore bulls (365.52 ± 39.19 kg) were randomly allocated to 24 pens (n = 4/pen), which were assigned to the treatments: (1) Control (no feed additives); (2) FO (500 mg/kg of DM); (3) MON (27 mg/kg of DM); and (4) MON + FO (27 and 500 mg/kg of DM, respectively). The FO used was a blend of castor oil acid and cashew nut shell liquid composed of cardanol (200 g/kg), ricinoleic acid (90 g/kg) and cardol (40 g/kg). Each treatment was replicated 6 times. The experimental diets were formulated according to the Large Ruminant Nutrition System (Fox et al., Reference Fox, Tedeschi, Tylutki, Russell, Van Amburgh, Chase, Pella and Overton2004) and are shown in Table 1. The step-up adaptation program lasted 16 days, where two adaptation diets containing 70 and 78% were fed for 7 and 9 days, respectively. The finishing diet containing 86% concentrate was fed for 89 days. Cattle were fed ad libitum 3 times/day at 08.00 h (30% of the total ration), 11.00 h (20% of the total ration) and 16.00 h (50% of the total ration), with free-choice access to water.
Feed ingredients and chemical composition of high-concentrate diets fed to Nellore yearling bulls (n = 24) during adaptation and finishing periods
DM, dry matter; NDF, neutral detergent fibre; peNDF, physically effective NDF.
a Supplement contained: Ca: 9.80%; P: 4.50%; Mg: 4.40%; K: 6.15%; Na: 11.45%; Cl: 6.60%; S: 4.00%; Co: 48.50 mg; Cu: 516.00 mg; Fe: 30.00 mg; Mn: 760.00 mg; Se: 9.00 mg; Zn: 2516.50 mg. Sodium monensin (Rumensin; Elanco Animal Health, São Paulo, Brazil) was added at 1323.53 mg/kg of supplement and functional oils (Essential; Oligo Basics, Cascavel, Brazil) was added at 24 509.80 mg/kg of supplement and offered to yearling bulls in the treatments.
b Estimated by equations according to Large Ruminant Nutrition System (Fox et al., Reference Fox, Tedeschi, Tylutki, Russell, Van Amburgh, Chase, Pella and Overton2004).
Feedlot performance and carcass traits
At the beginning of the experimental period, and every 28 days, bulls were withheld from feed for 16 h before body weight (BW) assessment. The final BW assessment was performed after 21 days of the previous one. The dry matter intake (DMI) was calculated daily and expressed in kilograms, and as a result, average daily gain (ADG) and gain to feed (G : F) ratio were calculated. According to Bevans et al. (Reference Bevans, Beauchemin, Schwartzkopf-Genswein, McKinnon and McAllister2005), the DMI variation was calculated as the difference between two consecutive days and expressed in kg and as a percentage of variation (DMI of previous day/DMI of current day × 100). In order to estimate the net energy for maintenance (NEm) and net energy for gain (NEg) of experimental diets, the methods described by Zinn and Shen (Reference Zinn and Shen1998) were used.
After 105 days, on feed cattle were transported 115 km (~3 h) to a commercial abattoir. Hot carcass weight (HCW) was obtained after kidney pelvic fat removal. The 12th rib fat thickness, Biceps femoris (BF) fat thickness, longissimus muscle (LM) area and marbling were measured via ultrasound at the beginning and the end of the study, following the method described by Perkins et al. (Reference Perkins, Green and Hamlin1992). Images were collected using an Aloka SSD-1100 Flexus RTU unit (Aloka Co. Ltd., Tokyo, Japan) with 17.2-cm, 3.5-MHz probe access.
Feeding behaviour and particle sorting
Cattle were submitted to visual observation of 24 h, every 5 minutes, to evaluate feeding behaviour. The visual observation was performed on day 53 of the study using a method adapted from Robles et al. (Reference Robles, González, Ferret, Manteca and Calsamiglia2007). Samples of diets and orts were collected for chemical analysis of dry matter (DM) and neutral detergent fibre (NDF; Van Soest et al., Reference Van Soest, Robertson and Lewis1991) to determine the intake of NDF and DM. The DMI and NDF intake per meal in kilograms was calculated by dividing DMI and NDF by the number of meals per day. Eating rate of DM, rumination rate of DM, eating rate of NDF and rumination rate of NDF were calculated according to Pereira et al. (Reference Pereira, Cruz, Arrigoni, Rigueiro, Silva, Carrara, Santos, Cursino and Millen2016). Samples of diets and orts were also collected for determination of particle-size distribution, which was performed by sieving using the Penn State Particle Size Separator and reported on an as-fed basis as described by Heinrichs (Reference Heinrichs1996). Selective consumption was determined as follows: n intake/n predicted intake, in which n = particle fraction screens of 19 mm (long), 8 mm (medium), 1.18 mm (short) and a pan (fine). Selective consumption values equal to 1 indicate no sorting, those <1 indicate selective refusals (sorting against) and those >1 indicate preferential consumption (sorting for).
Rumenitis score and rumen morphometrics
Rumenitis evaluation was performed after cattle evisceration, and all entire washed rumens were scored. Rumen epithelium was classified according to the incidence of lesions (rumenitis) and abnormalities (e.g., papillae clumped) as described by Bigham and McManus (Reference Bigham and McManus1975) using a scale of 0 (no lesions and abnormalities noted) to 10 (severe ulcerative lesions). Also, a 1-cm2 fragment of each rumen was collected from cranial sac and placed into a phosphate-buffered saline solution for future morphometric measurements according to Resende Júnior et al. (Reference Resende Júnior, Alonso and Pereira2006). Manually, the number of papillae per square centimetre of rumen wall (NOP) was determined; 12 papillae were randomly collected from each fragment and scanned and mean papillae area (MPA) was determined using an image analysis system (Image Tool, version 2.01 alpha 4, UTHSCSA Dental Diagnostic Science, San Antonio, TX). The rumen wall absorptive surface area (ASA) in square centimetres was calculated as follows: 1 + (NOP × MPA) − (NOP × 0.002), where 1 represents the 1 cm2 fragment collected and 0.002 is the estimated basal area of papillae in square centimetres.
Likewise, a 1-cm2 fragment of each rumen was collected from the ventral sac for histological assessment. Histological sections were stained with hematoxylin and eosin, embedded in paraffin wax and sectioned (Odongo et al., Reference Odongo, AlZahal, Lindinger, Duffield, Valdes, Terrell and McBride2006). Histological measurements, such as papillae height, papillae width, papillae surface area, keratinized layer thickness and the mitotic index, were determined in four papillae per animal using computer-aided light microscope image analysis. For the mitotic index, the number of cells exhibiting mitotic figures was determined using the same microscope as just described, and the final data were expressed as a percentage of 2000 cells.
Statistical analysis
The experimental design was a completely randomized block with a 2 × 2 factorial arrangement of treatments. Initial BW was utilized as a criterion for block formation. Pens were considered experimental unit for this study. Tests for normality and heterogeneity of treatment variances were performed before analysing the data. The data were analysed using the MIXED procedure of SAS (SAS Inst., Inc., Cary, NC) and Tukey test to compare means. The models included the main effects of FO (FO or no FO), MON (MON or no MON) and FO × MON interaction. For carcass traits variables, analyses of variance of final LM area, final 12th rib fat, final BF fat thickness and final marbling, included the initial measurement covariate when appropriate (P ⩽ 0.05), which was significant only for final LM area. Results were considered significant at the P < 0.05 level.
Results
Feedlot performance and carcass traits
Results of feedlot performance of Nellore yearling bulls fed either FO or MON are presented in Table 2. For the first 28 days on feed, no FO main effect (P > 0.05) was observed for any of the feedlot performance variables evaluated, except for DMI variation, where cattle fed FO had lower DMI variation expressed both in % (no FO: 12.66; FO: 10.56; P = 0.043) and in kg (no FO: 1.22; FO: 1.10; P = 0.049). Cattle fed MON presented lower final BW in kg (no MON: 408.22; MON: 400.90; P = 0.045), ADG in kg (no MON: 0.87; MON: 0.48; P = 0.003) and DMI in kg (no MON: 8.78; MON: 7.34; P < 0.01). Moreover, cattle consuming MON had poorer G : F ratio (no MON: 0.099; MON: 0.066; P = 0.022) and lesser DMI variation in kg (no MON: 1.24; MON: 1.08; P = 0.013).
Feedlot performance of Nellore yearling bulls (n = 24) consuming high-concentrate diets containing sodium monensin or functional oils
FO, functional oils; MON, sodium monensin; BW, body weight; ADG, average daily gain; DMI, dry matter intake; G : F, gain to feed; NEm, net energy for maintenance; NEg, net energy for gain; DM, dry matter. a,b,c Values within a row with different superscripts differ.
After 56 days on feed, no FO main effect (P > 0.05) was observed for most of the variables evaluated, except for DMI variation expressed in %, where cattle that had FO added to the diet presented lesser variation (no FO: 8.93; FO: 7.73; P = 0.049). Likewise, an interaction was observed (P = 0.023) between FO and MON for DMI variation expressed in kg, in which cattle receiving no feed additives had greater DMI variation when compared to those that were fed either MON or FO (Table 2). Cattle fed MON presented lower DMI in kg (no MON: 9.96; MON: 8.28; P < 0.01), however, no differences in terms of ADG (no MON: 1.15 kg; MON: 1.08 kg; P = 0.448) and G : F ratio (no MON: 0.116; MON: 0.130; P = 0.147) were observed when MON was added to the diets.
After 84 days on feed, no FO main effect (P > 0.05) was observed for any of the feedlot performance variables evaluated. An interaction was observed between FO and MON for DMI variation expressed both as % (P = 0.049) and kg (P = 0.036), where cattle on control group had greater DMI variation when compared with cattle receiving either MON or FO (Table 2). When the DMI variation was expressed in kg, cattle consuming MON or MON + FO had lesser variation than those receiving only FO. Cattle fed MON had lower DMI expressed in kg (no MON: 10.18; MON: 8.66; P < 0.01), which improved G : F ratio (no MON: 0.102; MON: 0.122; P < 0.01) without impacting ADG (no MON: 1.04 kg; MON: 1.06 kg; P = 0.837) and final BW (no MON: 471.1 kg; MON: 475.8 kg; P = 0.507) were observed.
Overall, after 105 days on feed, no FO main effect (P > 0.05) was observed for any of the feedlot performance variables evaluated. An interaction was observed between FO and MON for DMI variation expressed both as % (P = 0.048) and kg (P = 0.046), where cattle on control group had greater DMI variation when compared with cattle receiving either MON or FO (Table 2). When the DMI variation was expressed in kg, cattle consuming MON or MON + FO had lesser variation than those receiving only FO as well. Cattle fed MON had lower DMI expressed in kg (no MON: 10.28; MON: 8.87; P < 0.01), improved G : F ratio (no MON: 0.112; MON: 0.126; P = 0.038) and increased NEm (no MON: 1.6 Mcal/kg of DM; MON: 1.8 Mcal/kg of DM; P = 0.01) and NEg (no MON: 1.0 Mcal/kg of DM; MON: 1.1 Mcal/kg of DM; P = 0.01), but no differences were observed related to ADG (no MON: 1.15 kg; MON: 1.12 kg; P = 0.634) and final BW (no MON: 505.0 kg; MON: 503.7 kg; P = 0.872) were observed.
Results of carcass characteristics for Nellore yearling bulls fed either FO or MON are presented in Table 3. No FO main effect (P > 0.05) was observed for any of the carcass traits variables assessed. However, there was an interaction between FO and MON for final marbling (P = 0.030). Cattle fed either only FO or MON had greater final marbling when compared to those on the control group (Table 3). Cattle fed MON presented thinner final 12th rib fat (no MON: 4.26; MON: 3.41; P < 0.01), and final BF fat thickness expressed in mm (no MON: 5.84; MON: 4.73; P < 0.01), as well as slower rate of adipose tissue accretion as shown by 12th rib fat daily gain in mm (no MON: 0.020; MON: 0.013; P = 0.01) and BF fat daily gain in mm (no MON: 0.029; MON: 0.019; P < 0.01).
Carcass characteristics of Nellore yearling bulls (n = 24) consuming high-concentrate diets containing sodium monensin or functional oils
FO, functional oils; MON, sodium monensin; BF, back Fat measured on Biceps femoris muscle; LM, longissimus muscle. a,b Values within a row with different superscripts differ.
Feeding behaviour and particle sorting
Results of feeding behaviour of Nellore yearling bulls fed either FO or MON are presented on Table 4. No FO main effect (P > 0.05) was observed for any of the feeding behaviour variables evaluated. However, cattle fed MON had lower DMI (no MON: 10.19 kg; MON: 8.93 kg; P < 0.01), DMI per meal (no MON: 0.95 kg; MON: 0.71 kg; P = 0.01), NDF intake (no MON: 4.35 kg; MON: 3.65 kg; P < 0.01) and NDF intake per meal (no MON: 0.40 kg; MON: 0.29 kg; P < 0.01). In addition, cattle fed MON consumed both DM (no MON: 16.09 min/kg; MON: 19.93 min/kg; P = 0.041) and NDF (no MON: 37.03 min/kg; MON: 49.09 min/kg; P < 0.01) slower, as well as ruminated both DM (no MON: 32.57; MON: 36.81; P = 0.01) and NDF (no MON: 75.34; MON: 90.86; P < 0.01) slower than cattle not fed MON.
Feeding behaviour of Nellore yearling bulls (n = 24) consuming high-concentrate diets containing sodium monensin or functional oils
FO, functional oils; MON, sodium monensin; DMI, dry matter intake; ER, eating rate; DM, dry matter; RR, rumination rate; NDF, neutral detergent fibre. a,b Values within a row with different superscripts differ.
With respect to particle sorting, results are presented in Table 4. No FO or MON main effect (P > 0.05) was observed for the sorting of long particles. However, an interaction between FO and MON was observed for sorting of medium (P = 0.038), short (P = 0.041) and fine particles (P = 0.035), where cattle fed only FO consumed more medium and short particles, and less fine particles, than animals consuming only MON, MON + FO, or no feed additives.
Rumenitis score and rumen morphometrics
Results of rumenitis score and rumen morphometrics of Nellore yearling bulls fed either FO or MON are presented in Table 5. No FO main effect (P > 0.05) was observed for any of the variables evaluated, except for papillae width (no FO: 0.44 mm; FO: 0.41 mm; P = 0.05), where cattle receiving FO had smaller width. An interaction was observed for NOP (P = 0.035), in which cattle consuming only FO or MON had greater NOP than those receiving FO + MON; and for papillae area expressed as % of ASA (P = 0.039), where cattle fed any of the feed additives evaluated had a larger area than those on the control group. Cattle fed MON presented larger MPA (no MON: 0.48 cm2; MON: 0.60 cm2; P = 0.01), ASA (no MON: 31.06 cm2; MON: 37.23 cm2; P = 0.016), papillae surface area (no MON: 1.85 mm2; MON: 2.33 mm2; P = 0.01), as well as greater papillae height (no MON: 4.92 mm; MON: 5.73 mm; P = 0.046) and mitotic index (no MON: 1.64%; MON: 2.02%; P < 0.01).
Rumen morphometrics of Nellore yearling bulls (n = 24) consuming high-concentrate diets containing sodium monensin or functional oils
FO, functional oils; MON, sodium monensin; ASA, absorptive surface area; KLT, keratinized layer thickness. a,b Values within a row with different superscripts differ.
Discussion
The decreased DMI variation during the first 28 days on feed, which includes the adaptation period, presented by cattle receiving FO may contribute to decrease ruminal pH variation and increase production and absorption of short-chain fatty acids from the rumen (Meyer et al., Reference Meyer, Erickson, Klopfenstein, Greenquist, Luebbe, Williams and Engstrom2009); however, this decrease did not positively impact feedlot performance overall or even during the first 28 days. Typically, the DMI variation is greater for adapting than for adapted cattle, which was reported in this study, because during adaptation the intake still increases and adjusts to animal's requirements. According to Schwartzkopf-Genswein et al. (Reference Schwartzkopf-Genswein, Hickman, Shah, Krehbiel, Genswein, Silasi, Gibb, Crews and Mcallister2011), DMI variation equal to or less than 10%, probably does not cause a negative impact on the feedlot performance of cattle fed in individual pens. However, for group-fed cattle, the threshold for DMI variation should be lower than 10% due to the lower variance on DMI throughout the feeding period. Cruz et al. (Reference Cruz, Pereira, Millen, Arrigoni, Martins and Costa2016) conducted a meta-analysis on potential effects of natural DMI variation on feedlot performance and reported a negative effect on ADG when DMI variation went from 4 to 6% in pens containing three to five animals. In the present study, DMI variation across all treatments was below 10% but above 6%, which may explain why cattle consuming only FO sorted for medium and short, and against fine particles, which may have contributed to decreasing ruminal pH variation and led to the reduction in DMI variation.
Furthermore, the sorting against fine particles presented by cattle fed only FO may have decreased the amount of readily fermentable carbohydrates in the rumen, which may explain the absence of a positive effect on rumen morphometrics variables (except for papillae width) and feedlot performance when FO was added to the diet. According to Costa et al. (Reference Costa, Pereira, Melo, Resende Júnior and Chaves2008), propionate, which is an end product of readily fermentable carbohydrates and the main glucose precursor for ruminants, is also responsible for promoting the growth of metabolically active ruminal papillae. Based on the fact that feeding FO did not negatively impact rumenitis scores and rumen morphometrics, we cannot directly associate the sorting against fine particles shown by cattle fed FO with ruminal acidification. Therefore, reasons by which cattle receiving FO sorted against fine particles still remain unclear and deserve further investigation.
Despite the poorer feedlot performance during the first 28 days on feed, cattle fed MON improved G : F ratio throughout the study, which led to similar ADG, HCW and dressing percentage overall when compared to cattle not consuming MON. Duffield et al. (Reference Duffield, Merrill and Bagg2012) conducted a meta-analysis of the impact of MON on growing and finishing beef cattle bringing together 40 peer-reviewed articles and 24 additional trial reports. The authors reported that ADG increased by 2.5%, G : F ratio improved by 6.4% and DMI was reduced by 3%. In this study, feeding MON decreased DMI by 11.1%, had no negative effect on cattle ADG, and improved G : F ratio by 13.7% when compared to not feeding MON.
Cattle fed MON had lower DMI throughout the study due to the slower eating and rumination rates of DM and NDF (Table 3), which may have negatively impacted ruminal passage rate (Kp). Pereira et al. (Reference Pereira, Carrara, Silva, Silva, Watanabe, Tomaz, Arrigoni and Millen2015) also observed decreases on eating and rumination rates of DM and NDF when feedlot Nellore cattle consumed MON. Likewise, the negative effect on DMI caused by feeding MON may be explained by the increase of propionic acid in the rumen, which is the precursor of glucose and was previously reported to be negatively correlated with DMI in ruminants (Larsen et al., Reference Larsen, Relling, Reynolds and Kristensen2010). Furthermore, the negative effect on cellulose-degrading bacteria also slows the outflow of fibrous particles from the rumen, which negatively affects DMI. Linneen et al. (Reference Linneen, Harding, Smallwood, Horn, Jennings, Goad and Lalman2015) reported that when steers were supplemented with MON, particulate passage rate was reduced by approximately 42%. In general, adding MON into feedlot cattle diets decreases DMI to such an extent that it does not affect ADG, which results in improved G : F ratio (Duffield et al., Reference Duffield, Merrill and Bagg2012), as observed in the current study.
Some studies have not reported negative effects of feeding MON on carcass characteristics (Duffield et al., Reference Duffield, Merrill and Bagg2012). However, the slower rate of fat deposition observed in this study, which led to thinner final 12th rib and final BF fat, may be associated with the acetate : propionate ratio decrease (Ellis et al., Reference Ellis, Dijkstra, Bannink, Kebreab, Hook, Archibeque and France2012), which may negatively impact the lipid metabolism. Carbohydrates, such as cellulose and hemicellulose, when fermented in the rumen, produce large amounts of acetic acid, which is the precursor of carcass fat deposition. In the present study, despite the negative effects on fat deposition by adding MON into feedlot diets, it was not observed as a negative impact on HCW and dressing percentage. Despite the negative effects on subcutaneous fat deposition, cattle fed MON increased marbling, which is a result of increasing propionate and consequently glucose for the host. Smith and Crouse (Reference Smith and Crouse1984) reported that glucose is the major contributor to de novo fatty acid biosynthesis in intramuscular adipose tissue.
With respect to rumen morphometrics, cattle fed MON presented greater ruminal development, which may be associated with the increase of propionate. Ellis et al. (Reference Ellis, Dijkstra, Bannink, Kebreab, Hook, Archibeque and France2012) observed modification on volatile fat acids (VFA) production by increasing the propionate concentration and reducing the acetate concentration on beef cattle fed MON. Propionate is responsible for promoting the growth of metabolically active ruminal papillae (Costa et al., Reference Costa, Pereira, Melo, Resende Júnior and Chaves2008). Pereira et al. (Reference Pereira, Carrara, Silva, Pinto, Vicari, Pereira, Arrigoni and Millen2014) reported larger ASA in Nellore cattle fed high-starch diets containing MON. Since starch is one of the propionate precursors, the greater ASA presented by feedlot Nellore cattle fed MON in this study may have contributed to greater absorption of VFA, which may have led to an improved G : F ratio. The association of FO + MON reduced NOP when compared to the feeding of either FO or MON alone, similar to the effect reported on marbling for this study. It seems that the association of FO and MON in high-energy diets negatively impacts the propionate metabolism in the rumen; however, it deserves further investigation since the literature is still scarce on studies that combine feed additives to evaluate meat and rumen variables.
Thus, the adding of FO into high-concentrate diets decreased DMI variation throughout the feedlot period, especially in the first 28 days on feed. On the other hand, the adding of MON into high-concentrate diets improved feedlot performance and promoted greater development of rumen epithelium of Nellore cattle, as well as decreased DMI variation. Both feed additives, when added to the finishing diets, caused positive impacts on Nellore cattle, however in different magnitudes. Despite the positive effect on DMI variation, the feeding of FO was not as efficient as feeding MON to positively impact the performance of Nellore cattle fed high-concentrate diets.
We would like to thank São Paulo State Foundation (FAPESP) (Grant number: 2015/11927-3).
Financial support
The research was funded by São Paulo State Foundation (FAPESP) and Oligo Basics Agroindustrial.
Conflict of interest
The authors declare there are no conflicts of interest.
Ethical standards
All the procedures involving the use of animals in this study were in accordance with the guidelines established by the São Paulo State University Ethical Committee for Animal Research (protocol number 22/2015).
