Effects of Steam-Flaked Grains On Foals Growth Performance And Faecal Microbiota

There is little objective information concerning the effect of steam-aked grains on foal’s growth performance and faecal microbiota.To determine the effects of steam-aked grains on foal’s growth performance and faecal microbiota.Faecal samples were collection from 18 foals which had been fed corn, oat or barley diets over the 60 days of the experiment. Body weight and measurements were collected. Next-generation sequencing of the V3+V4 region of the 16S rRNA gene was used to assess the microbial composition of faeces. Alpha diversity, Venn graph, Relative abundance and beta diversity are presented.


Introduction
The gut microbiota performs essential roles in the maintenance of growth and health of animals (Marcio et al.,2015). In equine species, the gut bacterial microbiota is essential due to its role in cellulose fermentation and short-chain fatty acid production as the primary energy sources for growth and health (Glinsky et al.,1976). Equines are able to graze on high-bre, low-energy fodder due to the complex microbial community in their gut. However, factors such as diet (Daly et al.,2012), age, management, gut disease (Costa et al.,2012) and weaning have all been shown to cause changes in the gut microbiota of equines. Corn, oat, and barley are often included in equine diets to increase energy (Svihus et al.,2005).
They have similar amylose and amylopectins, but differ in the proportion of those polysaccharides, and also in the morphology of the starch granule (Tester et al.,2006). In the small intestine, the starch of grains is digested by amylolytic enzymes and absorbed as glucose. Any starch which is not digested in the small intestine of the equine will be delivered to the caecum and colon, and fermented by bacterial microbiota. In the hindgut, starch fermentation can lead to increased numbers of amylolytic bacteria, including lactobacilli and streptococci. This can increase the volatile fatty acid and lactic acid concentrations, decrease pH, and decrease the number of cellulolytic bacteria (Defombelle et al.,2001;Medina et al.,2002;Willing et al.,2009) To improve the e ciency of grain feeding to equines and to prevent diseases and health problems associated with the fermentation of starch in the hindgut, the digestibility of starch in the small intestine must be enhanced. Currently, the digestibility of starch in the small intestine is enhanced mainly through the selection and processing of grains. The purpose of grain processing is to enhance digestibility in the small intestine by changing its physical and chemical structure. Processing methods utilizing a combination of heat, moisture and pressure can disrupt the starch granule structure, destroy crystalline starch formations, increase the water solubility of starch and physically expose starch to digestive enzymes (Kienzle et al.,1997;Rowe et al.,1999). Steam-pressed tableting is the most commonly used process for starch. Generally, grains are processed by steam at 100-110℃ for 30 to 60 minutes. The grains are then ground into particles of speci c density by a pair of reverse rotating preheated rollers, which are dried and cooled to achieve safe water storage. Steampressed tablets made from grain improve the digestibility of starch in the small intestine. However, studies have not investigated how the source of the starch and processing affects equine growth and hindgut microbiota. We hypothesized that changes to the hindgut bacteria in response to dietary starch would be affected by the processing of starch. The objective was to compare the effects of adding oats, corn or wheat middlings to a forage-based diet on equine hindgut microbiota.

Animals
Eighteen healthy 5-month-old weaning Kazakh foals with a starting bodyweight of 112.36 ± 7.50 kg were studied. The foals were born in March, weaned in August, and the trial undertaken between August and October(60 days). Foals were selected from the same local pasture and were clinically normal with no history of systemic illness. Ivermectin was used for deworming the foals before weaning.
Foals were fed the same alfalfa, dry hay and concentrate supplement. Starch was introduced into the diet from steam-pressed corn, steam-pressed oats or steam-pressed barley. All foals were randomly allocated to one of the 3 treatments: corn group, oat group, or barley group (6 foals per group). The sum of starch in each starch source was used to adjust feed amounts to provide 2 g starch/kg bodyweight (Dry Matter,DM basis). The feed amount of the concentrate supplement was calculated according to the feed amount of the starch. The foal weighed 112 kg at 0-30 days and required 224 g of starch. The average weight of the 31-60 day old foals was 126 kg and the starch requirement was 252 g (Tables 1& 2).The nutrient levels of the foal's diet during the test period was shown in Table 3.

Management
During the 60 days study, the foals were housed in individual, partially covered runs with red brick (8×6 m). Each run contained an automatic water source and the feeding area of the run was equipped with red brick, and a large wood tub secured to the wall. Horses were allowed turnout per day in dry lot paddocks after feeding.
Weight and body measure data collection The body weight of foals was assessed using a loadometer (range 1000 kg), the body height and length were taken using a measuring rod, and the chest and shank circumference were taken using a tape measure (range 5 m). The data were collected before the rst feed in the morning on the rst, thirtieth, and sixtieth day of treatment.

Faecal sample collection
Faecal samples were collected by rectal palpation using one rectal sleeve per foal. Samples were stored in plastic sterile containers (RNA free) and snap frozen in liquid nitrogen before storage at -70℃ until DNA extraction. Samples were collected before the rst feed in the morning on the sixtieth day of treatment.

DNA extraction
Total genome DNA from samples was extracted using a CTAB/SDS method (Magoč et al.,2011). DNA concentration and purity was assessed on 1% agarose gels. DNA was diluted to 1 ng/µl using sterile water.
PCR ampli cation of the V3 + V4 region of 16S rRNA gene was undertaken. The primers used for the ampli cation were 341F: CCTAYGGGRBGCASCAG and 806R: GGACTACNNGGGTATCTAAT.
16S rRNA genes were ampli ed using the speci c primer with the barcode. Paired-end reads from the original DNA fragments are merged by using FLASH.
(1) Paired-end reads was assigned to each sample according to the unique barcodes. Sequences were analyzed using QIIME (Caporaso et

Data analysis
Weight and body measure data analysis Data were analyzed by Statistical Analysis System(SAS) (2013) as a randomized block design, considering the starch source as the main effect and the replicate as a block. The MIXED procedure was used to analyze body weight and body measurements. The variance structure adopts CS(Compound Symmetry), the data is the least square mean value, the signi cance judgment standard is P < 0.05, and the Ls means method is used for multiple comparison.
Alpha diversity and relative abundance analysis Statistical analyses on the microbial dataset were performed in R. Diversity and richness for both diets were evaluated using the Shannon diversity and Chao1 richness indices. T-test, LEfSe, and Anosim statistical analysis methods were used to test the signi cance of differences in the species composition and community structure of the group samples. Tax4Fun was used to predict and analyze the microbial community in the samples. To investigate microbial community structure of faecal in foals.

Growth performances
All foals remained in good health without diarrhea, stress or intestinal in ammation throughout the treatment. All foals adapted to the treatment diets and feeding management without problems. Table 4 shows the effect of the grain source on the growth performance of foals. The weight and body size leastsquare results of the foals during the experimental period showed that the body weight of the foals in the barley fed group was signi cantly higher than that of the oat fed group (P = 0.0152), and the effects of barley and corn supplementation on the body weight and body size of the foals were better than that of the oat group. However, after supplementing with corn, oats, or barley, weight and body size changed signi cantly as the foal aged (P < 0.01). There were no differences amongst the treatments in body height, body length, chest and shank circumference of foals (P > 0.05). The total and average daily increases in body weight and body size of the foals are shown in Fig. 1. The source of grains could effect the total and average daily increases in body weight and body size of the foals. The total weight gain and daily weight gain of the barley fed group was signi cantly higher than those foals fed corn or oats (P = 0.0185). There were no differences between the treatment groups in terms of body height, body length, chest or pipe circumference (P > 0.05).
Alpha Diversity was used to analyze the microbial community diversity within the faecal samples (Li et al.,2013). The single-sample diversity analysis (Alpha Diversity) could re ect the richness and diversity of the microbial community in the sample, including a series of statistical analysis indices to evaluate the differences in the species richness and diversity of the microbial community in each sample. Alpha diversity indices of the faecal bacteria are shown in Table 5. There were no signi cant differences in PD(Compute Faith's phylogenetic diversity metric)_whole_tree index, ace index, or goods coverage of all groups. There were signi cant differences in observed species index, Shannon index, Simpson index and Chao1 index of all groups. The observed species index was higher with the oat fed group than the corn fed group (P < 0.05). The Shannon and Simpson indices were higher in the barley fed group than in the corn or oat fed groups (P < 0.05, P < 0.05, P < 0.05 and P < 0.05). The Chao1 index was higher in the oat group than either the corn or barley group (P < 0.05 and P < 0.05). Venn graph based on operational taxonomic units The Venn diagram (Fig. 2) shows the three different feeds and the number of common OTUs between the groups, and the number without overlap which represent the unique OTUs of the group. The common OTUs of the three groups is 1745. The number of unique OTUs is 311 in the corn fed group, 383 in the oat fed group, and 356 in the barley fed group. The number of unique OTUs in the oat fed group is higher than in the corn or barley fed and suggests that more bacterial species were detected in the oat group.
Relative abundance of faecal bacteria

LDA Effect Size
LEfSe (LDA Effect Size) (Segata et al.,2011) is an analytical tool for the discovery and interpretation of high-dimensional biomarkers, which can be used for comparison between two or more groups. It emphasizes statistical signi cance and biological correlation, and can be used to nd statistically different biomarkers between groups. Figure 4 shows the characteristics of different abundance and associated categories of the results. LEfSe's statistical results include three parts, which are (1) the histogram of LDA value distribution, (2) the evolutionary branch (phylogenetic distribution), and (3) the abundance comparison of biomarkers with statistically signi cant difference among different groups. The histogram of LDA value distribution (Fig. 4A) shows that the statistically different biomarkers bacteria were Bacilli, Lactobacillales, Lactobacillaceae, Lactobacillus and Lactobacillus hayakitensis in the oat fed group; the Streptococcaceae, Streptococcus and Proteobacteria in the corn fed group; and the Clostridia, Clostridiales, Lachnospiraceae and Ruminococcaceae in the barley fed group. The evolutionary branch (phylogenetic distribution) and the abundance comparison of biomarkers showed that eight bacteria were Lactobacillaceae, Streptococcaceae, Lactobacillales, Bacilli, Lachnospiraceae, Ruminococcaceae, Clostridiales and Clostridia in all groups, respectively (Fig. 4B).

Function prediction of faecal bacteria
Tax4Fun (Aßhauer et al.,2015) is an R package based on 16S Silva database for predicting the function of intestinal and soil environmental samples. The prediction accuracy is high, and the experimental results show that the PICRUSt function prediction is better than the PICRUSt function prediction, especially suitable for soil and other complex environment samples. The function prediction of Tax4Fun was realized by using the nearest neighbor method based on the minimum 16S rRNA sequence similarity. Speci cally 16S rRNA gene sequences were extracted from the whole genome of prokaryotes in the KEGG database and compared to the SILVA SSU Ref NR database using BLASTN algorithm (BLAST Bitscore & GT;1500). The established correlation matrix corresponded to the whole genome function information of prokaryotes annotated through UProC and PAUDA to SILVA database, and realized the function annotation of SILVA database. The sequencing samples cluster OTU with the SILVA database sequence as the reference sequence, and then obtained the functional annotation information. Tax4Fun functional annotation clustering heat map of foal faecal samples are shown in Fig. 5. We detected functional information of 26 species of bacteria in foal faecal samples. There were 7 signi cantly increased species or functional information in the barley fed group: nitrogen xation, ark hydrogen oxidation, others, ark hydrogen oxidation, reductive acetogenesis, xylanolysis, methanogenesis by reduction of methyl compounds with H2, methylotrophy and aerobic chemoheterotrophy. In the corn fed group, we detected four species functional information which were signi cantly increased: aerobic chemoheterotrophy, cellulolysis, animal parasites or symbionts and nitrate reduction. There were three species functional information which were signi cantly increased in the oat fed group: methanogenesis by CO 2 reduction with H 2 , methanogenesis, and hydrogenotrophic methanogenesis.

Discussion
Starch is the storage form of plant nutrients. Grains contain a large amount of starch, which generally accounts for 20% ~ 40% of the dry matter of the diet. Most starches in grains are found in the endosperm, which contains about 70 %. Starch is a chain polymer composed of many glucose molecules linked together, divided into amylose and amylopectin. The average molecular weight of amylose was 1×10 5 to 9×10 6 (Dobladomaldonado et al.,2017), and the average molecular weight of amylose was 1×10 7 to 1×10 9 , being some of the largest polymers in nature (Tester et al.,2004). During the formation of starch granules, amylose penetrates among amylopectin, forming dense hydrogen bonds, preventing the invasion of digestive enzymes, and reducing starch fermentation (Gómez et al.,2016). The higher the ratio of amylose to amylopectin, the lower the digestibility of starch (Svihus et al.,2005). Therefore, an increase in amylose ratio will make starch degradation more di cult.
The steam sheet is a common heat treatment for grain. The process not only utilizing the grain, but also changes its crystal. The structure makes cereal starches easier to digest in the small intestine. Perez et al. (2010) believed that compared with crushing treatment, steam compression treatment could achieve complete utilizing and increase the utilizing rate of starch. Meyer et al.(1993) reported the effects of varying grain processing methods on the pre-caecal starch digestion of corn in horses. The results showed that whole corn, cracked corn, ground corn, and popped corn pre-caecal digestibility were 28.9%, 29.9%, 45.6% and 90.1%, respectively. A study by Mohsen et al.(2020) showed that the average daily gain and feed e ciency were improved and nal body weight was higher in steam-aked corn grain-fed calves.
In our study, we showed that 2 g starch/kg BW day (DM basis) could improve the body weight gain of foals. In addition to the different sources of starch, the composition and nutritional composition of other feeds were the same, as well as the feed processing method and the breeding and management conditions of foals. Thus, differences in dietary nutritional effects can be attributed to the source of starch. Among the three kinds of starches, the composition of starches is different, among which the proportion of amylose and amylopectin was different. A study showed that the ratio of amylose to amylopectin could effect growth performance and pork quality in pigs. Potter et al.(1992) suggested that the upper limit of starch digestion in the equine small intestine was between 3.5 and 4 g starch/kg body weight per feeding [24] . However, others have suggested that in order to limit starch bypass to the large intestine that starch intake should not exceed 2 g/kg body weight per meal (Meyer et al.,1995;Kienzle et al.,1992). In this study, no digestive abnormalities were observed in the foals, indicating that the increased starch content in this study met the needs of the foals. The total weight gain and daily weight gain of the barley fed group was signi cantly higher than that of the corn and oat fed groups. The uniform and stable release of glucose by the body is a necessary condition for the absorption and utilization of glucose in the small intestine to promote the growth and development potential of young animals. Only when the release of glucose in the intestine is consistent with the demand for glucose by the body tissues and organs can the performance potential of young animals be fully achieved. Results of Weurding et al.(2001), showed that a continuous, uniform and slow-release mode of glucose provided energy supply for the balanced intestinal tract of piglets, and allowed their growth to reach the optimal level. A study by Nasir et al.(2015) showed that low-quality or high-quality barley can fully replace wheat grain in diets for starter pigs and achieve equivalent or better growth performance. In this study, the signi cant weight gain in the barley fed group may be related to the speed of starch release from barley; releasing glucose more in line with the growth pattern of the foals.
The grain in the diet was used mainly to increase the energy requirements of the horse. Corn, oats and barley were the primary energy feed for horses, which contain similar amylose and amylopectin. However, due to the different proportion of starch polysaccharides and the different size of starch granules, their digestion and effects on the intestinal tract of horses are other (Kong et al.,2003;Svihus et al.,2005). Studies have shown that there were differences in the digestibility of grain particles in the small intestine of horses. Starch was not digested in the small intestine was degraded by microbial fermentation upon arrival in the small intestine (Defombelle et al.,2004).
The starch that is not digested by enzymes in the horse's small intestine is transported to the hind intestine for microbial fermentation. It has been estimated that the utilized energy of starch produced by post-intestinal microorganisms is 35% − 40% less than that absorbed in the form of glucose digested by enteroglycans (Kienzle et al.,1994;Black et al.,1971). Also, starches not digested by intestinal amylase arrive in the intestine. They are fermented by starch-breaking bacteria, causing fundamental changes in the structure of the posterior intestinal ora and increasing concentrations of volatile fatty acids and lactic acid in the caecum and colon (Garner et al.,1978).
Different starch sources and other intakes of starch would affect the number of starch decomposition bacteria. The amount of grain the foals were fed determined the amount of starch that they consumed.
When fed a low-content grain diet, the starch intake was low, and the number of starch decomposition bacteria increased in horses fed with corn and wheat, but not in horses fed with oats (Harlow et al.,2016).
When the amount of grain was high, the number of starch-breaking bacteria increased in horses fed with oats and corn, while the increase of starch-breaking bacteria was more signi cant in horses fed with a high corn diet. Such a massive difference in microbial digestion of corn starch and oat starch is directly related to the low sensitivity of corn starch to enzymic digestion in the small intestine of horses (Rosenfeld et al.,2008). As alpha-amylase in the small intestine of horses showed low activity in the digestion of corn starch, a large amount of undigested corn arrived in the intestine. This was fermented by microorganisms, increasing starch-breaking bacteria (Harlow et al.,2015). Therefore, the varying starch sources could explain the in uence of different starch sources on the microbial diversity of the hindgut, especially on the total starch decomposers. The starch source has been shown to affect the extent of the starch bypass to the equine hindgut (Radicke et al.,1991;Defombelle et al.,2004;Rosenfeld et al.,2008).
The starch source would also affect changes to the gastrointestinal microbial community that are induced when the grain is added to a forage-based diet. Ren et al.(2019) used 16S rRNA sequencing technology to reveal the potential mechanism of steam-pressed corn to improve the production performance of ruminants, and fed steam-pressed corn to cows. They subsequently found Firmicutes and mutants in the rumen microbial community. The relative abundance of the Proteobacteria tended to increase or increased signi cantly, succinic acid vibrios (Succinivibrio) and rothia (Roseburia) and slaughter bacteria genera (Blautia) the relative abundance of the starch decomposition microbes such as add, reduce the relative abundance of cellulose decomposition microbes.
In this study, alpha diversity indices showed that fed grains could change the structure and diversity of bacteria in foal's faecal matter. The sample from barley fed foals signi cantly increased Shannon and Simpson indexes and altered the community structure of bacteria in foal's faecal matter. The sample from oat fed foals had a signi cantly increased Chao1 index and changed the diversity of bacteria in the foal's faecal matter. The relative abundance of faecal bacteria results showed that Actinobacteria, S treptococcaceae, Lactobacillaceae, Streptococcus, Lactobacillus and Actinobacillus were signi cantly increased when corn was fed. The relative abundance of Tenericutes, Rikenellaceae and Lactobacillus in faecal matter was signi cantly increased when oat was provided. Lachnospiraceae and Agathobacter were signi cantly increased when barley was fed. A study by Defombelle et al.(2001) looked at the effect of changing the diet from 100% hay (100:0) to 50% hay and 50% barley. The results showed that concentrations of lactate-utilizing bacteria were not signi cantly altered in the caecum and colon of ponies. However, the concentrations of lactobacilli and streptococci were signi cantly increased 5 hours after altering the diet, and then signi cantly decreased after 29 hours.
Ruminococcaceae bacteria have previously been identi ed as brinolytic bacteria (Daly et al.,2012). In this study, the LEfSe result showed that Ruminococcaceae amounts were signi cantly different in faecal matter from barley fed foals. A study by Bulmer et al.(2019) demonstrated that even a small addition of starch to the diet was enough to reduce this bacterial Ruminococcaceae population. Our research showed that the relative abundance of Ruminococcaceae in the corn, oat and barley group was: Ruminococcus_sp_HUN007 (0.16%,0.13% 0.47%, respectively) and Ruminococcus avefaciens (0.01%, 0.02%, 0.17%, respectively). Research has shown that the extent of the alterations to faecal microbiota with the addition of starch to the diet can be in uenced by the source of starch feed (Harlow et al.,2016).
Our study showed that Lactobacillales, Lactobacillaceae, Lactobacillus and Lactobacillus hayakitensis (14.19%, 18.09% and 7.11% in faecal matter when fed corn, oat and barley, respectively) were signi cantly different in the oat fed group compared to the corn and barley fed group. An increase in lactic-acid producing bacteria has also been reported to be coupled with a corresponding decrease in brolytic bacterial abundance (Daly et al.,2012). The extent of starch digestion in the small intestine can be affected by processing (Potter et al.,1992;Meyer et al.,1995), so the amount of starch that will result in bypass to the large intestine is likely to vary with both starch source and processing. In this study, the grains were treated by steam pressure, and the processing method was consistent. The reason for the difference in bacterial ora may be related to the structure and proportion of cereal starch depending on the source.
The results of functional prediction showed that long-term feeding of three different grains had varying effects on digestive physiology and health of foals. Results from Bulmer et al.(2019) showed that the dietary change resulted in alterations in behaviour and faecal microbiota. The diet could change faecal microbial community composition and relative abundance. Increased starch in the current study had an undesirable effect on behaviour and gut microbiota; it made the ponies more reactive in their behaviour and moved the microbial community composition of the gut towards dysbiosis. However, the opposite was true of the high bre diet. In this study, we didn't observe the abnormal behaviour and faecal bacteria in foals when fed corn, oat or barley for 60 days. Intestinal microorganisms have speci c effects on immune function, nutrient absorption, and even enzyme metabolism (Martin et al.,2010). The Tax4Fun prediction was that the diet of corn, oat or barley could affect digestion and metabolism in foals, and the effects of the three grains supplementation on microbial function prediction were different.

Conclusions
In our study, differences were seen in the faecal microbiota of foals fed either corn, oat or barley, and also differences in the overall growth of the foals. Different grains have different impact on faecal microbiota, which are mainly related to the grain sources. Further investigation is required to look at the potential impact of changes in the microbiota on the functional impact on foals when fed grains.

Declarations Ethics declarations
Ethics approval and consent to participate All protocols were approved by theAnimal Care and Use Committee of Xinjiang Agricultural University (permission number 2018012).All methods were carried out in accordance with relevant guidelines and regulations for the use of animal subjects. The study was carried out in compliance with the ARRIVE guidelines.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article.

Competing interests
The authors declare that they have no competing interests.

Funding
The study was supported by grants from National Natural Science Foundation of China (31860649).

Authors' contributions
Li X B, Yang K L, Zang C J. were involved in study design and execution, data interpretation, and manuscript preparation. Huang X X, Ma C, Chen K X, Zhao G D, Li X Y, Zhang W J, Li Q. performed animals feed, sample collection and data analysis. All authors read and approved the nal manuscript.