1. Introduction
Consumer apprehension related to antibiotic use within production agriculture has inspired ruminant nutritionists to find alternative feed-additive strategies to support production animals’ overall health and performance. Over the past twenty years, probiotics have greatly expanded within livestock production. In 1989, the FDA strictly defined the use of probiotics in livestock production as “direct fed microbials” (DFMs), which were naturally occurring microorganisms that altered ruminal fermentation and intestinal function [
1,
2].
The ruminant scientific community and industry leaders have started to show increasing support for DFMs. These additives support normal gut functions, and consistent results [
3,
4,
5,
6,
7,
8] have demonstrated health support and improved nutrient digestibility and performance for beef and dairy cattle.
A range of DFMs have been developed to enhance immune response, increase feed efficiency, and improve carcass characteristics in beef cattle [
9,
10,
11]. Most DFMs on the market are a combination of multiple bacteria strains; such combination is used to elicit the most significant response via numerous modes of action. The most common bacterial strains in cattle diets are lactate-producing and lactate-utilizing bacteria. This combination is used due to their synergistic effects on ruminal fermentation [
12]. Lactate-producing bacteria, such as
Lactobacillus animalis, have shown resilience within acidic conditions, allowing the ruminal microbiota to adjust to the environment. One benefit of lactate production in the rumen is stimulating lactate-utilizing bacteria, such as
Propionibacterium freudenreichii [
1,
10,
13]. By utilizing lactate,
Propionibacterium spp. produces propionate, the primary glucose precursor in ruminants [
14]; therefore, the combination of both strains allows the conversion of lactate into propionate, which favors the rumen’s energetic efficiency while maintaining a more stable microbial ruminal environment [
15]. Moreover, numerous studies have demonstrated growth performance improvements and immunological benefits when this combination of bacteria was fed to beef cattle [
3,
10,
16,
17,
18,
19].
In addition to
Lactobacillus animalis and
Propionibacterium freudenreichii, spore-forming bacteria such as Bacilli have become a popular DFM choice. Bacilli have various beneficial characteristics, such as germination and proliferation in the rumen, providing long-lasting benefits within the entire gastrointestinal tract [
20]. Species such as
Bacillus subtilis have demonstrated enhanced growth of proteolytic bacteria because of greater nitrogen digestibility and increasing ammonia concentrations [
21,
22,
23]. More specifically,
Bacillus subtilis is reportedly capable of maintaining beneficial bacterial populations throughout the GI tract [
24]; therefore, its use can promote beneficial, durable changes to the microbiota within the intestine [
25], recovery from diarrhea [
26], and improvements in average daily gain and feed efficiency [
27,
28,
29].
Other Bacillus species, such as
Bacillus licheniformis, have been found to produce specific enzymes, such as amylase and cellulase [
30], leading to improved starch [
23] and fiber digestibility [
22]. Hence, both Bacilli species have demonstrated improvements in the digestibility of dry matter and neutral detergent fiber in various forages [
6,
31] while simultaneously improving starch digestion. [
31].
Considering the individual benefits found within these unique bacteria for the gastrointestinal tracts of ruminants, a novel DFM containing
L. animalis 506,
P. freudenreichii 507,
B. licheniformis 809, and
B. subtilis 597 (BOVAMINE DEFEND
® Plus) has been developed with favorable in vitro and in vivo effects on the health and performance of beef cattle [
7,
8]. We hypothesize that feeding a multispecies bacterial DFM would improve growth performance and carcass weights. Therefore, our objective was to evaluate the effects of feeding a multispecies bacterial DFM on the growth performance, carcass characteristics, and health variables of finishing-feedlot beef cattle.
2. Materials and Methods
2.1. Study Population
The study began in August 2022 at a commercial cattle feedlot (Hy-Plains Feedyard, LLC) near Montezuma, KS, USA. Before the study’s initiation, the Veterinary Research and Consulting Services, LLC, and Institutional Animal Care and Use Committee (IACUC number 1014) approved all animal procedures.
A total of 1625 black- and black–white-faced steers (average body weight [BW] = 371 ± 8.4 kg) were sourced from grazing stocker operations in New Mexico and California and enrolled in the study from August to September 2022. Steers were age- and source-verified and part of a non-hormone-treated cattle (NHTC) branded-beef program. Upon arrival at the feedlot, steers were unloaded and housed in pens by origin and provided ad libitum access to prairie hay and water. Before initial processing, a rest period of 4 to 6 days was allowed. During the rest period, steers had access to the starter ration (without DFM) and ad libitum access to prairie hay and water.
2.2. Arrival Processing
Before processing, all steers were evaluated by a licensed veterinarian to identify any chronic animals to be removed from the study. All steers were processed following the same protocol. Serially numbered ear tags, color-coded for each pen, were matched to the low-frequency electronic identification tags. Modified live virus vaccine (Pyramid® 3, Boehringer Ingelheim, Duluth, GA, USA) containing Infectious Bovine Rhinotracheitis Virus, Bovine Viral Diarrhea Virus (types 1 and 2), and Bovine Respiratory Syncytial Virus was administered subcutaneously (SC). Ivermectin (1.0 mL/50 kg of body weight; Ivermax® 1%, Aspen Veterinary Resources, Liberty, MO, USA) was administered SC. Oxfendazole (1.0 mL/50 kg of body weight; Synanthic®, Boehringer Ingelheim Animal Health USA, Inc., Duluth, GA, USA) was administered orally.
All products were administered in accordance with Beef Quality Assurance guidelines. After processing, all steers were weighed by pen in drafts on a certified platform scale to determine pen enrollment weight.
2.2.1. Treatment Allocation
At arrival processing, steers were sorted into three optimal marketing groups using a chute-side technological program to optimize sorting decisions (PenPoint, Elanco Animal Health, Greenfield, IN, USA) over each day of enrollment. Within each marketing group, steers were assigned to 1 of 2 treatment groups, resulting in 6 pens enrolled per day and a total of 24 pens across the study:
- (1)
Negative control—(CON; no DFM feeding)
- (2)
Direct-fed microbial—feeding of 50 mg/head per day of a DFM containing L. animalis 506, P. freudenreichii 507, B. licheniformis 809, and B. subtilis 597 (BDP; BOVAMINE DEFEND® Plus; Chr Hansen, Milwaukee, WI, USA).
According to the marketing group allocation outcome, steers were assigned treatment groups according to the order in which they entered the chute. Treatment groups were randomly selected from sorting, with the initial treatment group drawn being assigned to the first group within a marketing cluster and the subsequent treatment group allocated to the second group in the same marketing cluster. This process was repeated thrice daily, and a new sequence was randomly chosen at the start of each block.
2.2.2. Feed, Housing, and Water
Upon arrival, steers were offered long-stemmed prairie hay with starter ration delivered on top (1% of BW, DM basis,
Table 1). Steers were stepped up onto the finisher ration over 21 days and were fed twice daily using a slick bunk feeding program. Feed was delivered twice daily in commercial Roto-mix feed delivery trucks at 0800 h and 1300 h, and bunks were evaluated each morning at 0630 h for feed refusals. Diet samples were collected from bunks weekly and frozen at −20 °C. Samples were composited at the conclusion of the experiment and submitted to a commercial laboratory (Servitech Laboratories, Dodge City, KS, USA) for analysis of DM, CP, NDF, ADF, Ca, P, and S (
Table 1). Both pens within each block were transitioned to the next ration on the same day, and prairie hay was available (free choice) to each pen during each ration transition. Monensin (Rumensin
TM, Elanco Animal Health, Greenfield, IN, USA) was fed for improved feed efficiency, and tylosin (Tylan™, Elanco Animal Health, Greenfield, IN, USA) was fed to reduce liver abscesses. No additional feed-grade antibiotics were fed throughout the trial.
The two pens within each block were housed in adjacent pens, and all pens were within the same area of the feedlot. Water was provided ad libitum through an automatic float-activated system shared between pens. An average of 68 steers (range: 55 to 85) were enrolled per pen, with an average enrollment weight of 371 kg (range: 324 to 418 kg) across all blocks. Pen area per steer averaged 26.5 m2 (range: 20 to 32 m2), and bunk space averaged 32 cm (range: 24 to 38 cm) per steer enrolled. Pen metrics were similar (within one steer) for each experimental block.
2.2.3. Interim Weight
Both pens within a block had an interim pen weight collected on the same day at an average of 82 days on feed (DOF; range: 72 to 88 DOF). Steers were weighed in drafts by pen for interim body weight on certified platform scales (WH. Scale Co., Topeka, KS, USA).
2.2.4. Harvest
Steers were harvested based upon visual estimation of adequate finish, feed intake, and cattle supply availability for the NHTC branded-beef program. The average DOF at the time of harvest was 133 (range 106 to 151 days). Steers were fed the morning of the shipment and shipped in late afternoon, and both pens within a block were shipped on the same day. Steers were weighed in drafts on a certified platform scale by pen prior to being loaded on commercial trucks with a 5% shrink applied to final live weight, gain, and feed conversion calculations. Cattle arrived at a commercial packing plant in Arkansas City, KS, USA (375 km) to begin harvest the following morning (fasting period of approximately 8 h). All steers were harvested from 5 December 2022 to 2 February 2023 and shipped on 5 dates (2–8 blocks shipped per date).
2.2.5. Carcass Outcomes and Liver Scoring
At the harvest facility, trained personnel from West Texas A&M University—Beef Carcass Research Center cross-referenced individual animal identification tags with plant carcass identification. Carcass quality grade was provided by a manual USDA grader at the plant, and yield grade was generated on an individual carcass basis based on the processor’s visual camera grading system. Carcasses that did not have a camera score recorded (n = 34; CON = 20 and BDP = 14) were assigned a yield grade of 3 if the quality grade was USDA Choice or better or assigned a yield grade of 2 if the quality was USDA Select or USDA Standard. Personnel from Beef Carcass Research Center, blinded to the treatment group, scored livers based on the Elanco Liver Scoring System (Elanco Animal Health, Greenfield, IN, USA). Edible livers without any abnormalities were classified as normal while livers with 1 or 2 small abscesses or inactive scars were classified as A−; livers with 1 to 2 large abscesses or multiple small abscesses were classified as A; and the A+ score was used to describe multiple large abscesses present. Ruptured abscesses and those with tissue adhesions were categorized as A+ liver scores. The presence of liver flukes, telangiectasis, congestive heart failure, and cirrhosis were also recorded and categorized as “other” under liver abnormalities. No samples were collected for culture or histopathology evaluation.
2.2.6. Economic Analysis
Cost of gain was calculated on a deads-in and deads-out basis. Briefly, deads-in used initial body weight and head days from mortalities and removals were included in the calculation whereas deads-out used the initial body weight and head days from mortalities and removals were excluded from the calculation. Cost of gain was determined by dividing the total costs for the pen (feed costs, medicine costs, processing costs, and yardage) by the total kilograms of weight gained during the feeding period. Commercial prices were included in the cost of BDP feeding. All other feed ingredients used during the study were included in feedlot closeout expenses. Sale weight had a 5% shrink applied to all final weights, and head days were used to determine performance outcomes.
2.2.7. Statistical Analysis
Data were evaluated using a commercial software program (R Studio Team 2023, Version 4.2, Boston, MA, USA). Pen served as the experimental unit for all outcomes comparing treatment groups. Continuous outcomes (body weight, average daily gain [ADG], feed to gain [F:G], dry matter intake, cost of gain, carcass weight, and dressing percent) were evaluated as a randomized complete-block design with linear mixed models. Binomial outcomes (BRD first and second treatment, BRD mortality, digestive mortality, other mortality, overall mortality, removals, and total outs (deads + removals)) were evaluated using generalized mixed models. Health and performance outcomes were assessed at time of interim weight and closeout. Health data were also evaluated from interim weight to closeout. All models included a fixed effect of treatment group and random effect of block. Differences having p ≤ 0.05 were considered statistically significant, with tendencies described as 0.05 < p ≤ 0.10. Descriptive cumulative mortality and total outs were evaluated by treatment group and DOF. Multinomial cumulative link mixed models were used to evaluate distribution outcomes by treatment group for quality grade, yield grade, and liver scores. Multinomial models included fixed effects of treatment group and random effects of block and repeated measures of the lot. Pairwise comparisons were performed within each category across treatment groups if the overall distribution was significantly (p ≤ 0.05) different.
3. Results
Initial BWs were not different (
p = 0.92) among the treatments (370 ± 8.4 kg) at the beginning of the experiment (
Table 2). Steers in the BDP group tended to have a heavier body weight (
p = 0.06; 530.0 vs. 533.6 kg for CON vs. BDP, respectively), greater ADG (both deads-in and deads-out;
p = 0.06), and lower F:G conversion deads-in (
p = 0.06; 5.39 vs. 5.19; CON vs. BDP, respectively) when compared to the CON group. Steers fed BDP had significantly (
p = 0.05; 5.26 vs. 5.09; NC vs. BDP, respectively) improved F:G deads-out through the interim weight period compared to those fed CON. No differences due to treatment in morbidity or mortality were identified at the time of interim weight measurement.
There was a tendency for fewer total deads (
p = 0.07; 0.72 vs. 0.10%) and a significant reduction in the number of total outs (
p = 0.01; 1.61 vs. 0.11%) in the BDP group compared to the CON group from the time of interim weight checking to closeout (
Table 3). A total of nine steers were removed from the study and not marketed with the cohort (BDP = 1; CON = 8). Causes for removals included BRD (CON = 4), not qualifying for the NHTC program (BDP = 1; CON = 1), being non-performing (CON = 2), and being musculoskeletal (CON = 1). Steers in the BDP group had fewer removals (
p < 0.05) and total outs (deads and removals;
p < 0.01) compared to the CON group (
Table 4). The cumulative mortality and total outs are shown in
Figure 1 and
Figure 2.
Performance outcomes at closeout (
Table 5) showed that the F:G deads-in basis was significantly (
p = 0.05; 7.08 vs. 6.71) improved for the BDP group compared to the CON group. The average daily gain on the deads-in basis (
p = 0.06), F:G deads-out basis, and cost of gain for both deads-in (
p = 0.06; 159.83 vs. 151.87), and deads-out (
p = 0.08; 155.02 vs. 150.20) tended to be improved for the BDP group compared to the CON group.
There were no significant differences (
p > 0.36) in carcass characteristics (
Table 6), indicating that the improved growth rate and feed efficiency outcomes did not impact carcass traits. The distribution of liver abscess scores was not different between the treatment groups (
p = 0.52), but total liver abscesses were reduced in the BDP group compared to the CON group (
p = 0.01).
4. Discussion
A 6% greater overall ADG for steers offered BDP corroborates with the findings reported by Galyean et al. [
32] and Swinney-Floyd et al. [
33], which supplemented
Lactobacillus acidophilus (1 × 10
8 CFU/animal-daily) and
Propionibacterium strain P63 (1 × 10
9 CFU/animal-daily) to cattle fed a 90% concentrate diet, wherein researchers found a 2–3% increase in feed efficiency. Dry matter intake was not affected by the supplementation of the current DFM, which corroborates results found in several experiments using similar DFM combinations [
9,
16,
17,
18,
32,
34]. The increased ADG and similar DMI levels resulted in an improved F:G of approximately 6% for animals offered BDP. Previous studies evaluating the supplementation of
Lactobacillus acidophilus and
Propionibacterium freudenreichii showed 2% to 7.3% improvements in F:G [
16,
19,
32]. When cattle were offered a 90% concentrate diet (65% steam-flaked corn-based diet) with a DFM containing
Lactobacillus (
L. acidophilus; 1 × 10
9 CFU/animal-daily), a tendency for a 5% improvement in F:G was observed for cattle offered the DFM [
9]. Lawrence et al. [
35] evaluated the effects of
Lactobacillus animalis (1 × 10
9 CFU/animal-daily) and
Propionibacterium freudenreichii (2 × 10
9 CFU/animal-daily) on dairy cattle performance. They reported high rates of mastitis in the herd, which may have masked any DFM effect. Although the literature results are not directly comparable due to differences in bacterial species, strains, concentrations, and diet types, it seems reasonable to assume that supplementing current bacterial DFM mixtures to cattle does not seem to cause adverse effects. In contrast, potential improvements in cattle growth performance have been repeatedly reported. Krehbiel et al. [
10] suggested that using a combination of lactate-producing bacteria and lactate-utilizing bacteria improved the ADG (2.5% to 5%). At the same time, other performance variables, such as DMI and F:G, are less consistent across DFM studies. This highlights the importance of designing experiments to measure variables other than growth performance, in which potential mechanisms of action, such as nutrient digestion and ruminal morphology, can be elucidated.
Similarly, the utilization of Bacilli as a DFM for cattle has been justified by its spore-forming and stability features [
36], as well as its enzyme production [
37]. Recently, in vitro DM and NDF digestibility have improved after incubating
B. licheniformis and
B. subtilis with different forage-based substrates [
31,
38]. The various types and amounts of fibrolytic enzymes produced by the different Bacilli strains may explain these improvements [
37,
39]. Moreover, in vitro, starch degradation has been enhanced in high-starch concentrates incubated with the same
Bacillus spp. [
31]. This indicates that the broad range of enzymes produced may lead to enhanced nutrient utilization and a rumen fermentation profile necessary to improve beef cattle performance [
5,
40]. Therefore, the different modes of action that Bacilli present may complement the efficacy of
L. animalis and
P. freudenreichii in enhancing the performance and health of feedlot cattle, as reported herein.
Hot carcass weight was not affected by the treatments. Galyean et al. [
32] offered
L. acidophilus and
P. freudenreichii at various concentrations, and HCW increased on average by 2.2%. However, Vasconcelos et al. [
18] used DFM treatments like those reported by Galyean et al. [
32] and did not observe differences in HCW or other carcass characteristics. Due to cattle availability, the current study population comprised NHTC, and these were harvested earlier than was optimal based on carcass weights and yield grades. The dressing percentage, longissimus muscle area, liver scores, and USDA quality grades were not affected by treatment, like in the data reported by Galyean et al. [
32], Brashears et al. [
9], and Vasconcelos et al. [
18].
The modulation of rumen pH using a DFM may be a potential reason for the decrease in total liver abscess as a reduction in the number of total outs through closeout for the current study. The time when the ruminal pH of feedlot cattle is below 5.6 (subacute rumen acidosis) is associated with increased liver abscesses [
41,
42]. Liver abscesses are primarily believed to be secondary to rumen acidosis, which allows bacteria to access the portal vein [
43]; damage to the epithelial lining of the gastrointestinal tract can be a causative agent. In addition to pH, another causative agent of liver abscesses,
Salmonella enterica, is found throughout the gastrointestinal system [
44,
45]. Rumen pH was not monitored in the current study, and tylosin was administered to both treatment groups to control liver abscesses. Additional research is needed to determine the repeatability of feeding
Lactobacillus animalis,
Propionibacterium freudenreichii,
Bacillus licheniformis, and
Bacillus subtilis for mitigating liver abscesses.
Direct-fed microbials have been shown to improve cattle health outcomes [
46,
47]. Studies with DFMs have focused on binding undesirable Gram-negative pathogens, such as Escherichia coli O157:H7, from the intestinal tract [
9]. Feeding DFMs may improve an animal’s immune response, but the response magnitude depends on the species and bacterial strain [
47,
48]. A yeast product with the primary mechanism of action occurring in the small intestine reduced the BRD first treatment by 28.4% and severe A+ liver abscesses [
49]. The current study did not identify a difference in morbidity, but the cattle had a low incidence of BRD morbidity. The total-outs separation primarily occurred later in the feeding period. The cattle removed later in the feeding period had most of the feeding costs incurred throughout the feeding period. The authors hypothesize that the reduction in the number of total outs may be attributed to improved gastrointestinal health from the DFM fed in this study. Additional research is warranted to evaluate the health findings further.
Limitations of the current study include that the study cattle were NHTC and harvested earlier than the industry standard; however, the study cattle yield was still greater than 10% yield (grades 4 and 5). An interim weight was collected to mimic when a re-implant would occur for traditional feedlot cattle. The authors anticipate a similar magnitude of response in cattle administered an implant and fed for a more extended period. Still, additional research is needed to support or refute the hypothesis.