A new approach: preventive protocols with yeast products and essential oils can reduce the in-feed use of antibiotics in growing-finishing pigs

ABSTRACT The objective of this study was to evaluate the effects of yeast products (YP) and essential oils (EO) in total or partial replacement to in-feed antibiotic protocols (growth promoter and prophylactic), both in recommended doses and in overdose of prophylactic antibiotics (PA), on growth performance, and diarrhea incidence in the growing-finishing pigs; and fecal microbiota in market hogs. Four hundred pigs (20.36 ± 2.64 kg) were assigned to five treatments in a randomized block design: diets with prophylactic and growth promoter antibiotics (ANT); ANT with 30% more PA (ANT+30); diets with less PA and YP (ANT+Y); diets with less PA, YP and EO (ANT+Y+EO); and antibiotics-free diets with YP and EO (Y+EO). The content of the active components of the YP was 60% purified β-1,3/1,6-glucans extracted from Saccharomyces cerevisiae yeast (Macrogard), 20% functional water-soluble MOS (HyperGen), and 18% MOS, extracted from Saccharomyces cerevisiae yeast (ActiveMOS). From 0 to 14 d, pigs of the ANT+30, ANT+Y, and ANT+Y+EO treatments showed a greater body weight (BW) and average daily gain (ADG) compared to pigs from the Y+EO group. From 14 to 35 d, pigs of ANT+30 and ANT+Y+EO treatments were heavier than Y+EO group. At 105 d, ANT pigs had a higher BW than the Y+EO group. For the entire period, ADG of ANT pigs was greater, and feed conversion ratio better than Y+EO pigs. From 0 to 35 d, pigs of the Y+EO treatment showed a higher diarrhea incidence compared to pigs of the other groups. From 49 to 70 d, ANT+Y and ANT+Y+EO treatments showed a lower diarrhea incidence than Y+EO group, which remained the case during the overall period. At 105 d, the alpha diversity of fecal microbiota by Shannon Entropy was lower in ANT, ANT+30, and Y+EO groups than observed for ANT+Y+EO group. The abundance of Firmicutes phylum and Firmicutes/Bacteroidetes ratio was higher in ANT than in ANT+Y+EO pigs. Proteobacteria phylum abundance in ANT+Y+EO was higher than ANT, ANT+Y, and Y+EO. Peptostreptococcaceae family abundance was higher in ANT, ANT+30, and ANT+Y groups than in ANT+Y+EO and Y+EO groups. ANT+Y+EO and Y+EO groups show a lower abundance of SMB53 genus than ANT and ANT+30 groups. In conclusion, the use of YP and EO, in partial replacement to the in-feed antibiotic protocols, does not reduce the growth performance, can replace antibiotic growth promotors, and reduce the in-feed use of PA in growing-finishing pigs. The use of YP and EO, together with PA, increases the microbial diversity, despite having important genera for weight gain in less abundance. Overdose of PA does not improve growth performance and reduces microbial diversity, which does not characterize it as an efficient preventive protocol.


INTRODUCTION
In-feed antibiotics have been widely used to increase growth rates and prevent diseases in pigs (Van Boeckel et al., 2015), as much in the postweaning period as in the growing and finishing phases (Gaskins et al., 2002).However, routine use of antibiotics, and the overdose mainly, in food-producing animals has significantly contributed to the increasing emergence of multidrug resistant pathogens, incurring a major health concern in both animals and humans (Landers et al., 2012;Tang et al., 2017).There is currently a great interest in the production of pigs without antibiotics (Liu et al., 2018), which has been a reality for the European Union since 2006 (Dewulf et al., 2022).Therefore, to avoid the negative effects of removing antibiotics from the diets of pigs, changes in management and nutritional strategies may be required (Kil and Stein, 2010), as well as the use of validated alternative additives (Heo et al., 2013;Liu et al., 2018).
Yeast products (YP) are proposed as alternatives to antibiotics in the livestock industry (Burdick Sanchez et al., 2021).Dietary supplementation of YP has been paid increasing attention for improving immune function and intestinal development in swine (Broadway et al., 2015;Xu et al., 2018;Zhaxi et al., 2020).β-glucan, a functional polysaccharide of d-glucose monomers linked by β-glycosidic bonds (Stier et al., 2014), can modulate the immune system and stimulate a cascade of pathways that enhance both innate and adaptive immune responses (Vannucci et al., 2013), besides promotion on intestinal function (Xiong et al., 2015).Mannan oligosaccharides (MOS) prevent the adhesion of pathogenic bacteria to intestinal epithelial cells by attachment to the mannose-binding proteins expressed on the bacterial fimbriae (Kogan and Kocher, 2007;Spring et al., 2015).In addition, MOS supplementation evidenced additional beneficial properties such as decreased incidence of diarrhea (Zhao et al., 2012;Valpotić et al., 2016;Song et al., 2019) and higher growth performance (Miguel et al., 2004;Agazzi et al., 2020).
Essential oils (EO) are promising blends as novel antibacterial agents that can be used in pig production.Their bioactive compounds derived have immune, antioxidative, and antimicrobial properties (Sharifi-Rad et al. 2017).EO enhance digestibility (Liu et al., 2012;Chitprasert and Sutaphanit, 2014), reduce the production of cells and molecules involved in the immune response (Brenes and Roura, 2010;Liu et al., 2013), and promote gut health by minimizing the effect of the pathogenic bacteria (Chitprasert and Sutaphanit, 2014;Zhang et al., 2020).The modulation of gut microbiota, resulting in improvement of the growth performance, is also observed in piglets supplemented with EO (Li et al., 2018;Omonijo et al., 2018), therefore is so studied as alternative to antibiotics (Gong et al., 2013;Zhang et al., 2020).Studies that associate YP with EO are not found, as well as protocols with different inclusions of antimicrobial additives.In addition, partial replacement of prophylactic antibiotics (PA) with additive protocols has not yet been demonstrated.We understand that there is a potential synergism in this association that can contribute beneficially to the growth performance of pigs and reduce the total use of antibiotics (growth promoter and prophylactic forms).
The involvement of gut microbiota in host metabolism and health is well accepted (Lynch and Pedersen, 2016), but some specific roles remain to be studied.The intestinal microbiota plays crucial functions in nutrient digestion and absorption (Backhed et al., 2015), the development of the host immune (Postler and Ghosh, 2017), the differentiation of intestinal epithelium (Sommer and Backhed, 2013), and the maintenance of intestinal mucosal barrier (Garrett et al., 2010).There is also a possible link between the intestinal microbiota and growth performance, mainly feed efficiency (FE) in pigs (Yang et al., 2016;McCormack et al., 2017;Xiao et al., 2017;Tan et al., 2018).Therefore, the association of performance tests with microbiome technology in market hogs treated with different preventive protocols can provide a solid contribution to the understanding of the microbiota and host metabolism relationship.
We hypothesized that replacing antibiotic growth promoters (AGP) with nutritional additives or reducing the frequency of use of PA, due to the inclusion of these additives, do not decrease the growth performance of pigs in the growingfinishing phase and modify the fecal microbiota in market hogs.Therefore, the objective of this study was to evaluate the effects of YP and EO in total or partial replacement to the in-feed antibiotic protocols (growth promoter and prophylactic), both in recommended doses and in overdose of PA, on growth performance, in the total consumption of antibiotics during the experimental period, diarrhea incidence in the growing-finishing pigs, and fecal microbiota in market hogs.

Animals, Experimental Design, and Housing
The experimental design and procedures were approved by the Ethics Committee on Animal Use of University of São Paulo under Protocol number 2172090321.The experiment was conducted in the growing-finishing facilities of the Animalnutri Research Center located within a commercial pig farm in Patos de Minas, Brazil.A total of 200 barrows and 200 gilts (DanBred sows and LQ1250 sires) with an average initial body weight of 20.36 ± 2.64 kg (63 d of age) were randomly allocated in a randomized block design (sex and initial body weight were the blocking factor).The experiment had five treatments: diets with prophylactic and growth promoter antibiotics (ANT); ANT with 30% more prophylactics antibiotics (ANT+30); diets with less prophylactics antibiotics and YP as prophylactic and growth promoter (ANT+Y); diets with less prophylactics antibiotics, YP, and EO as prophylactic and growth promoter (ANT+Y+EO); and antibiotics-free diets with YP and EO as prophylactics and growth promoter (Y+EO).The treatments are further detailed in Table 1.Eight replicates (10 pigs/pen) were used in the trial.The PA were composed of 200 ppm tiamulin and 440 ppm amoxicillin, and the overdose was 30% higher PA.The AGP used was 10 and 5 ppm enramycin in growing and finishing phases, respectively.The content of the active components of the yeast products 1 was 60% purified β-1,3/1,6-glucans extracted from Saccharomyces cerevisiae yeast (Macrogard, Biorigin, São Paulo, Brazil).The content of the active components of the yeast products 2 was 20% functional water-soluble MOS (HyperGen, Biorigin, São Paulo, Brazil).The content of the active components of the yeast products 3 was 18% MOS, extracted from Saccharomyces cerevisiae yeast (ActiveMOS, Biorigin, São Paulo, Brazil).The contents of the active components of the EO were a blend of 12% carvacrol and 6% cinnamaldehyde, capsaicin, anethole, and cineole (Activo, GRASP, Curitiba, Brazil).
The animals were housed in pens allowed a floor space of 1.40 m² per pig and had a partially slatted floor.All piglets were provided with feed and water in a five-space feeder (semiautomatic) and nipple drinkers.

Diets and Experimental Procedures
The experimental period was 105 d and it used six diets, all pigs were fed the same basal diet.The basal diet was formulated to meet or exceed the nutritional specifications suggested by Rostagno et al. (2017) for pigs during the growing to finishing phases (Supplementary Table S1) and the additives were added according to each treatment.The pigs had ad libitum access to feed and water throughout the experimental period.The basal diet did not contain growth promoter additives.
Feed intake per pen and individual body weights were recorded at 0, 14, 35, 49, 70, 84, and 105 d.Based on these data, the ADG, ADFI, feed conversion ratio (FCR), and the total consumption of antibiotics were calculated.For fecal scoring, the feces present in the pen were assessed daily and graded as normal feces (no diarrhea) or liquid or pasty stools (presence of diarrhea), following the method of Casey et al. (2007).At the end of the trial, the occurrence of diarrhea was calculated as percentage of the phase.

Sample Collections
On day 105 of the trial, fecal samples were collected from the six average-weight replicates (pens), totaling 36 samples.The three pigs with the closest individual ADG to the group ADG mean were chosen.Fecal samples were collected by digital rectal stimulation and a pool with these three samples was formed.All fecal samples were immediately frozen in liquid nitrogen and stored at −80 °C until analyses.

Fecal Microbiota Analyses
Total bacterial DNA was extracted from the fecal sample contents by using a commercial kit (ZR Fecal DNA MiniPrep kit; Zymo Research Corp., Irvine, CA, USA) according to the manufacturer's instructions.The extracted DNA was quantified by spectrophotometry at 260 nm using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., Wilmington, DE, USA).To assess the integrity of the extracted DNA, all samples underwent electrophoresis in 1% agarose gel, were stained with a 1% ethidium bromide solution, and visualized with ultraviolet light in a transilluminator.
Thereafter, the variable V4 region of the 16S rRNA gene was amplified using the universal primers 515F and 806R and KlenTaq Master Mix (Sigma).Amplification controls without a template were employed.The PCR conditions used were: 94 °C for 3 min (1 cycle), 94 °C for 45 s/50 °C for 30 s/68 °C for 60 s (18 cycles), and a last step of 72 °C for 10 min.The amplicons were quantified with Qubit using an HS dsDNA kit (Invitrogen), diluted to 500 pM, and pooled.Then, 16 pM of pooled DNA were sequenced using MiSeq reagent 500V2 (Degnan and Ochman, 2012).Sequencing was performed using an Illumina MiSeq sequencer (Illumina) obtaining paired-end reads of 250 bp as described.
For the sequences obtained, the data filtering was completed by removing low-quality base.The sequences after trimming were analyzed with the QIIME pipeline (Caporaso et al., 2010), including the extraction of operational taxonomic units (OTUs) and overlapping analyses of OTUs.OTUs were clustered with a 97% similarity threshold.To compare the sequences, the 2017 update (SILVA 128) of the SILVA ribosomal sequences database (Yilmaz et al., 2014) was used.To generate the classification of bacterial communities by identifying OTUs, 22,110 reads per sample were used.This was in order to normalize the data and not compare samples with a different number of reads, thus avoiding a sample bias.

Nucleotide Sequence Accession Number
All the raw sequences obtained after assembling and filtering were submitted to the NCBI site under accession number BioSample: PRJNA721984.

Statistical Analyses
The pen was used as the experimental unit for the statistical analysis of growth performance and antibiotics consumption, and the pooled samples were used as the experimental unit for the microbiome analyses.The data were analyzed using the SAS software statistical package 9.3 (SAS, Cary, NC), except for sequencing data.The Shapiro-Wilk test was used to evaluate parametric data distribution.If the data did not have a normal distribution, data transformation was performed using PROC RANK.The effects were analyzed using the SAS MIXED procedure appropriate for a randomized block design (initial weight and sex).When the F test (P < 0.050) showed a significant difference, Tukey's test was used to compare the means, with a significance level of 0.050.To analyze the diarrhea incidence, a generalized linear model (binomial analysis) was performed using the GENMOD procedure of SAS 9.3, with a significance level of 0.050.DNA sequencing was performed using the statistical metagenomics program STAMP: Statistical Analysis of Metagenomic Profiles.To compare the abundance of the genders identified between treatments, ANOVA (P < 0.050) with Tukey-Kramer post hoc test and Benjamini-Hochberg FDR multiple test correction was used.Only statistically different results were shown.The averages for biodiversity between treatments were compared using the number of OTUs and the Kruskal-Wallis test (P < 0.050), because they presented a nonparametric distribution according

Growth Performance
The growth performance results of the pigs during the experimental period are presented in Table 2.In the growing I phase (0 to 14 d), the pigs of the ANT+30, ANT+Y, and ANT+Y+EO treatments showed a greater BW (P = 0.001) and ADG (P = 0.001) compared to the pigs from the Y+EO group.The pigs of ANT+Y+EO group had better ADG than ANT pigs.In the growing II phase (14 to 35 d), the pigs of the ANT+30 and ANT+Y+EO treatments were heavier than Y+EO group (P = 0.010).In the growing III and IV, and finishing I phases, there were no differences between treatments for the performance variables.In the last phase (finishing II, 84 to 105 d), the ANT pigs had a higher BW than the Y+EO group (P = 0.023), but similar results to ANT+30, ANT+Y, and ANT+Y+EO treatments.In this phase, the pigs of ANT+30 and Y+EO groups had lower ADG than ANT pigs (P = 0.007).When taking into account the entire experimental period, the ADG of the ANT pigs was greater than Y+EO pigs but similar results to the ANT+30, ANT+Y, and ANT+Y+EO groups (P = 0.025).The pigs of ANT group had better FCR than Y+EO pigs and similar to other treatments (P = 0.011).

Antibiotics Consumption
As shown in Table 3, the experimental treatments provided significant variations regarding the amount of antibiotic consumed in each of the groups during the evaluation of the entire period.The pigs of the ANT+30 treatment showed the highest consumption of tiamulin (P < 0.001), amoxicillin (P < 0.001), and total antibiotics (P < 0.001).The ANT treatment obtained a higher consumption of tiamulin, amoxicillin, and total antibiotics than the ANT+Y and ANT+Y+EO groups, and lower when compared to the ANT+30 treatment.There were no differences between the ANT+Y and ANT+Y+EO treatments for all variables.Pigs from ANT and ANT+30 treatments showed no difference in enramycin consumption.As it is an antibiotic-free treatment, the Y+EO group does not show antibiotic consumption values, as well as the ANT+Y and ANT+Y+EO groups when analyzing enramycin.

Diarrhea Incidence
As shown in Fig. 1, during the growing phases I (0 to 14 d) and II (14 to 35 d), the pigs of the Y+EO treatment showed a higher diarrhea incidence compared to the pigs of the other groups (P < 0.001).In growing phase III (35 to 49 d), the ANT, ANT+30, and ANT+Y+EO treatments had a lower diarrhea incidence than the pigs of Y+EO group (P = 0.019).From 49 to 70 d (growing phase IV), the ANT+Y and ANT+Y+EO treatments showed a lower diarrhea incidence than the Y+EO     Yeast products and essential oils for pigs 7 group (P = 0.005), which remained the case during the overall period (P < 0.001).In the finishing I and II phases, there were no differences between treatments for diarrhea incidence.Considering the entire period, the ANT+Y+EO group had a lower diarrhea incidence than the ANT+Y group.The treatments had no effects on the diarrhea incidence from the 70th to 105th day of the trial (P = 0.631).

Fecal Microbiota
Regarding the biodiversity indicators of the bacterial communities, alpha diversity by Shannon Entropy were lower in the ANT, ANT+30, and Y+EO groups than observed for the ANT+Y+EO group (P < 0.050), but similar between other groups (Fig. 2).The most abundant phyla on fecal samples were Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, and Spirochaetes (Fig. 3).Firmicutes phylum and F/B ratio were higher abundant in the ANT group than in the ANT+Y+EO group (P < 0.050, Fig. 4).Proteobacteria phylum relative abundance in ANT+Y+EO was similar to ANT+30 and higher than ANT, ANT+Y, and Y+EO (P < 0.050, Fig. 5).
At lower taxonomic levels representing Firmicutes phyla, Peptostreptococcaceae family (a member of Clostridia Class) relative abundance was higher in ANT+30, ANT, and ANT+Y groups than in the ANT+Y+EO and Y+EO groups (P < 0.050, Fig. 6).ANT+Y+EO and Y+EO groups show a lower abundance (%) of SMB53 genus (Clostridiaceae family) than ANT and ANT+30 groups.Y+EO had shown the lowest levels between all groups (P < 0.010; Fig. 7).Alphaproteobacteria class abundance was higher in the ANT+30 group than any other group (P < 0.050.Fig. 8).

DISCUSSION
In-feed additives are used as long as they fulfill a variety of purposes.It is important that they are resistant to the processes used in in-feed manufacturing.In addition, they must be safe for animals and must not affect food safety.Those products should also share a positive image toward the public, who are increasingly sensitive to the way livestock production works  ( Gallois et 2009).Finally, these alternatives must be effective in their purpose, act as growth promoters and provide health benefits to pig.However, it is important to note that there is no single alternative to substitute in-feed antibiotics, and a combination of different alternatives to antibiotics may be the most promising method to reduce or replace antibiotics in animal feeds (Xu et al., 2021).
In this study, in the first two phases of evaluation, the replacement of antibiotics by additives maintained the BW and ADG of the pigs.In addition, the use of overdose promoted a greater BW than the use of additives alone.In the entire period, the total substitution of antibiotics by additives decreased the growth performance of pigs, but the reduction in PA use as well as the replacement of AGP by YP and EO additives did not interfere with the growth performance of pigs.The data regarding the consumption of antibiotics per treatment, followed as expected, so that the experimental groups with the replacement and reduction of the use of antimicrobials by the additives generated a lower intake of the three drugs used.This factor impacted the other variables.In the nursery phase, it is common to observe the replacement of antibiotics by additives without impairing growth performance (Xiong et al., 2015;Kiarie et al., 2018;Soler et al., 2018;Resende et al., 2020).This is mainly due to the improvement in intestinal morphology or digestibility (Long et al., 2018).Modulation of the microbiota is also widely observed (Resende et al., 2020).The changes are associated with the great challenges that weaning provides, as it is associated with intestinal inflammation and digestive disorders (Smith et al., 2010;Campbell et al., 2013;Moeser et al., 2017).In the growing-finishing phase, less results are found and can be inconsistent (Wierup, 2001).Cheng et al. (2018) confirmed the hypothesis that the use of EO as an alternative to antibiotics can improve intestinal epithelial absorptive function, gut microbiota composition, and antioxidative stress capacity in growing-finishing pigs, as well as Yan et al. (2011) when using an herb extract mixture.Xu et al. (2021) showed, through a meta-analysis, that plant extracts improve ADG and FCR for growing pigs and can be used as antibiotics substitute.The antimicrobial and anti-inflammatory properties of these additives change the microbiota and regulates intestinal permeability contributing to their antidiarrheal properties (Wang et al., 2017).
We observed a lower diarrhea incidence when EO were used in the protocol associated with antibiotics and YP, which contributes to these findings.However, some studies that evaluate the inclusion of additives, such as EO, plant extracts, or YP, in antibiotics-free diets, have not observed an effect on the growth performance of pigs (Davis et al., 2002;Janz et  Yeast products and essential oils for pigs al., 2007;Yan et al., 2010;Lowell et al., 2018).This inconsistency may be attributed to the different additives levels and animals used in each study (Yan et al., 2012) or differences of optimum concentration, purity, molecular weight, conformation, chemical modification, and solubility of YP in the diet formulation (Luo et al., 2019).Moreover, in this phase, the pigs may have a more developed digestive system, improved immunity, and increased resistance to intestinal disorders as the pig become older (Yan et al., 2010).
The pigs from ANT and ANT+30 groups did not show differences in growth performance in any of the phases, even though there was a difference of 22.83% for the total consumption of antibiotics between treatments, which shows that overdose antibiotics can often be ineffective.Resistance to antibiotics is an inherent side effect associated with the overuse, abuse, or substantial use of ANT (Williams-Nguyen et al., 2016;Manyi-Loh et al., 2018;Albernaz-Gonçalves et al., 2021).When resistant strains are extracted from their host, they are disseminated into the environment, and increase the spread of resistant genes among bacterial populations (Lin et al., 2015).The diverse range of novel antibiotic-resistance genes could be accessible to clinically relevant bacteria and play a critical role in the emergence of antibiotic resistance among pathogens (Pehrsson et al., 2013).Besides being ineffective, overdose antibiotics negatively affect the alpha diversity, reducing the total genus content, compared with the replacement of AGP and reduction of antibiotics.Elimination of antibiotics in diet shows a similar effect.Reduced alpha diversity is related to a less mature and stable microbiota, with a reduction of functional redundancy (Chen et al., 2017;Trevisi et al., 2021).The loss of microbial diversity is referred to as gut dysbiosis (Wilkins et al., 2019) and the use of antibiotics is associated with intestinal microbiota imbalance (Stecher et al., 2013;Jo et al., 2021) and reduced diversity (Knecht et al., 2014).Just like there are still persistent longterm impacts on the intestinal microbiota that remain posttreatment (Cecilia et al., 2007;Yin et al., 2015), our studies showed that antibiotics overuse also strongly impacts the diversity.
The supply of the YP, for pigs that received antibiotics during the growing-finishing phase, did not modify the alpha diversity indicators, differently of the use of YP and EO.Changes in alpha diversity were also not observed by Xu et al. (2018) when supplementing weaned piglets with yeast probiotics.The effect of YP on the immune system has been investigated in many studies (Broadway et al., 2015), but the results in the enteric microbiota can be contradictory.In addition, an overall increasing trend in alpha community diversity of the gut microbiome is observed during aging (Wang et al., 2019), suggesting that the swine gut microbiome matures after the finishing phase, and this can contribute to the equivalence between the treatments that included the YP or not.Zhang et al. (2020) and Li et al. (2018) also found no differences in alpha diversity in their studies with EO for weaned piglets.As the increase in microbial diversity in this study was found with the addition of both additives, we can observe a synergism between them, favoring the intestinal health these pigs.Microbiome diversity is used as a good indicator of gut health in humans (Menni et al., 2017) and animals (Hildebrand et al., 2013).
The most abundant phyla in fecal samples were Firmicutes and Bacteroidetes.The F/B ratio was highly abundant in ANT group than in ANT+Y+EO group.It is usual that the F/B ratio in swine gut microbiota gradually increases over time (Kim et al., 2012(Kim et al., , 2015)).Thus, ANT-treated pigs showed an accelerated increase in the F/B ratio (Kim et al., 2016).Pigs from ANT group received antibiotics the entire experimental period, unlike animals from the ANT+Y+EO group, promoting a 34.33% reduction in the total consumption of antibiotics by animals.A lower abundance of Firmicutes levels reflects on lower F/B ratios indicating a microbiome that favors gram-negative bacteria such as Bacteroides, Alistipes, Parabacteroides, and Prevotella (Stojanov et al., 2020).
Peptostreptococcaceae family (Firmicutes phylum) relative abundance was also higher when the antibiotics were used.The abundance of species of four families of the Firmicutes phylum (Streptococcaceae, Peptococcaceae, Peptostreptococcaceae, and Clostridiaceae) correlated positively with host weight gain (Kim et al., 2016).Firmicutes species metabolize available energy sources more effectively than Bacteroidetes species and consequently promote weight gain (Kallus and Brandt, 2012).Thus, manipulating gut microbial communities could control fat storage in pigs (Guo et al., 2008) and affect growth traits in pigs through host-microbe interactions (Lu et al., 2018;Oh et al., 2020).
ANT+Y+EO and Y+EO groups showed a lower abundance of Clostridium-related SMB53 genera than ANT and ANT+30 groups.Pigs from Y+EO group, which did not receive antibiotics at any phase, had shown the lowest levels among all groups.Gao et al. (2018) found functional analysis during the weaning and nursery periods that showed a convergent presence of carbohydrate metabolism, amino acid metabolism, DNA replication and repair, and membrane transportation and the genera.The SMB53 genus belongs to the Clostridiaceae family.Most members of this family have the ability to consume mucus-and plant-derived saccharides, such as glucose (Wuest et al., 2011).These findings are interesting because the pigs from Y+EO group showed lower growth performance during the evaluation, which may have happened due to poor use of feed ingredients.The fecal microbiota in high-feed efficiency pigs have a greater capacity to degrade dietary cellulose, polysaccharide, and protein and may have a greater abundance of microbes to promote intestinal health (Quan et al., 2019).In this study, the use of ANT seems to favor the microbiota for this purpose.
Greater FE increases profitability while reducing the environmental impact of pig production (Quan et al., 2020).This is especially important given that pig is one of the major sources of animal protein in human diet.The reduction in the use of antibiotics proposed in our study, and its effects on Yeast products and essential oils for pigs growth performance, fulfilled these assumptions and becomes a viable option.Furthermore, total withdrawal be assessed as a possible protocol when requested, without causing major losses in productivity.In recent years, analyzing the pig microbiota has gained interest because it allows for the prediction of the functional and metabolic capacity of such communities, which are believed to impact all aspects of host physiology including nutrient processing, energy harvesting, and growth performance (Tan et al., 2017;Quan et al., 2019).Thus, a better understanding of how the additives interfere with the finishing pig's microbiota is fundamental for the studies in the future since preventive protocols without antibiotics will be more and more demanded.

CONCLUSIONS
In conclusion, our results indicate that the use of YP and EO as growth promoters and as prophylactic, in partial replacement to the in-feed antibiotic protocols, does not affect the growth performance, can replace antibiotic growth promotors, and reduces the use of PA in-feed in growing-finishing pigs.In addition, the use of YP and EO increased the microbial diversity, despite having important genera for weight gain in less abundance.Overdose of PA does not improve growth performance and reduces microbial diversity, which does not characterize it as an efficient preventive protocol.
indicate significant differences between groups according to Tukey's test, P < 0.050.Data are expressed as means (eight replicates/treatment).4Relativeamount of antibiotic use between groups with ANT treatment as a reference value of 100%.

Figure 1 .
Figure1.Effect of experimental diets on the diarrhea incidence in pigs from growing to finishing phase.ANT: basal diet with antibiotics; ANT+30: basal diet with antibiotics in overdose; ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; Y+EO: basal diet with yeast products and essential oils.Data are expressed as means (eight replicates/treatment) and SEM is represented by vertical bars.*Significant differences between groups according to binomial analysis, P < 0.050.

Figure 2 .
Figure 2. Effect of experimental diets on the alpha diversity by Shannon Entropy in market hogs.ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT: basal diet with antibiotics; Y+EO: basal diet with yeast products and essential oils.Piglet is an experimental unit; six piglets/treatment.

Figure 3 .
Figure 3.Effect of experimental diets on the phylum levels in market hogs.Y+EO: basal diet with yeast products and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; ANT: basal diet with antibiotics; ANT+Y: basal diet with antibiotics and yeast products.The phyla present differ significantly in their abundance between treatments according to the Kruskal-Wallis test (P < 0.050).Piglet is an experimental unit; six piglets/treatment.

Figure 4 .
Figure 4. Effect of experimental diets on the Firmicutes/Bacteroidetes (F/B) ratio in market hogs.ANT: basal diet with antibiotics; ANT+30: basal diet with antibiotics in overdose; ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; Y+EO: basal diet with yeast products and essential oils.The F/B ratio means between different groups was compared by one-way ANOVA (P < 0.050) with Tukey's multiple comparisons test.Piglet is an experimental unit; six piglets/treatment.

Figure 5 .
Figure 5.Effect of experimental diets on Proteobacteria phylum relative abundance in market hogs.(A) Box plot showing the distribution in the proportion of Proteobacteria assigned to samples from treatments.(B) Post hoc plot for Proteobacteria indicating.ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT: basal diet with antibiotics; Y+EO: basal diet with yeast products and essential oils.Piglet is an experimental unit; six piglets/treatment.

Figure 6 .
Figure 6.Effect of experimental diets on Peptostreptococcaceae family relative abundance in market hogs.(A) Box plot showing the distribution in the proportion of Peptostreptococcaceae assigned to samples from treatments.(B) Post hoc plot for Peptostreptococcaceae indicating.ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT: basal diet with antibiotics; Y+EO: basal diet with yeast products and essential oils.Piglet is an experimental unit; six piglets/treatment.

Figure 7 .
Figure 7. Effect of experimental diets on SMB53 genus (Clostridiaceae family) relative abundance in market hogs.(A) Box plot showing the distribution in the proportion of SMB53 genus assigned to samples from treatments.(B) Post hoc plot for SMB53 genus indicating.ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT: basal diet with antibiotics; Y+EO: basal diet with yeast products and essential oils.Piglet is an experimental unit; six piglets/treatment.

Figure 8 .
Figure 8.Effect of experimental diets on Alphaproteobacteria class relative abundance in market hogs.(A) Box plot showing the distribution in the proportion of Alphaproteobacteria genus assigned to samples from treatments.(B) Post hoc plot for Alphaproteobacteria genus indicating.ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products, and essential oils; ANT+30: basal diet with antibiotics in overdose; ANT: basal diet with antibiotics; Y+EO: basal diet with yeast products and essential oils.Piglet is an experimental unit; six piglets/treatment.

Table 2 .
Effects of experimental diets on growth performance in pigs from growing to finishing phase 1 BW, body weight; ADG, average daily gain; ADFI, average daily feed intake; FCR, feed conversion ratio; SEM, standard error of the mean.2ANT: basal diet with antibiotics; ANT+30: basal diet with antibiotics in overdose; ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products and essential oils; Y+EO: basal diet with yeast products and essential oils.

Table 3 .
Effects of experimental diets on antibiotics consumption (g) in pigs from growing to finishing phase 1ANT: basal diet with antibiotics; ANT+30: basal diet with antibiotics in overdose; ANT+Y: basal diet with antibiotics and yeast products; ANT+Y+EO: basal diet with antibiotics, yeast products and essential oils; Y+EO: basal diet with yeast products and essential oils.