Beneficial effects of a calf starter versus forage on rumen development and bacteria populations in beef calves. Beneficial effects of a calf starter versus forage on rumen development and bacteria populations in beef calves

: Rumen development depends on the intake of solid food that is fermented into volatile fatty acids that stimulate the development of the rumen papillae in calves. The starter feeding can promote the growth of papillae in the rumen and as a consequence an earlier weaning. We evaluated the effects of calf starter on ruminal development, and productive response of lactating bull calves raised for meat in the tropics. Twelve male Brahman × Swiss American cross beef calves from a dual-purpose system were randomly assigned two treatments with six animals per treatment: milk-fed calves + Taiwan grass (Pennisetum purpureum, MT) and MT + calf starter, (MTS). Feed intake and growth were measured at 7-day intervals throughout until 210 d of age. At 90 days old, three calves from each treatment were harvested, and fluid and ruminal tissues were collected from the cranial, ventral, dorsal, and dorsal blind ruminal sacs for measurements of many papillaes per cm 2 (NP), papillae length (LP) and papillae width (WP). Ruminal bacterial genotype identification was determined by amplicon generation with the Illumina platform. Calf starter-improved weight (Live weight, LW) and average weight gain (ADG) and NP, but, LP and WP was similar in both treatments (p < 0.05). In calves with starter feed treatment, we observed the bacteria Desulfonauticus autotrophicus sp. nov.that was not previously reported in ruminants. Use of calf starter showed benefit for calves with improved feed intake and rumen development because promoted a greater number of rumen papillae.


INTRODUCTION
Dual-purpose livestock represent 78% of cattle production in the tropical regions of Latin America, and it is estimated that most of the calves destined for beef production come from this system (SOLORIO et al., 2016). The weaning weights of Brahman and Nellore beef calves, fed without milk restriction, are between 163 at 188 kg (MEDINA et al., 2005;HERNÁNDEZ et al., 2015); however, the weaning weight of beef calves from a dualpurpose system are lower due to milk restriction Medina et al. (SEGURA et al., 2017). This approach limits rumen development and functionality; at weaning, calves have fully undeveloped the capacity to use nutrients from forage, which affects post-weaning weight gain (SIMEONE & BERETTA, 2016).Rumen epithelium development is affected by the early intake of calf starter because it is fermented in the rumen, producing volatile fatty acids such as propionate and butyrate (KHAN et al., 2016); the latter functions locally in the ruminal wall to stimulate its development (KLEVENHUSEN et al., 2013;XIE et al., 2014). Calves that consume sodium butyrate through the calf starter or milk substitute have longer and wider ruminal papillae, and the weight of the reticulorumen tissue increases (SLUSARCZYK et al., 2010;KATO et al., 2011;WILSON et al., 2012). In addition, sodium butyrate regulates the proliferation, differentiation, and functionality of ruminal epithelial cells (SERBESTER et al., 2014). The calf starter must provide the elements required for the establishment and activity of ruminal microbial species to maximize the production of propionate and butyrate, and favor the development of the ruminal epithelium (DRACKLEY, 2008); facilitating an earlier weaning age, without affecting the productive response postweaning. High-starch diets affect microbial colonization and ruminal development (KHAN et al., 2008;PLAINZIER et al., 2012); therefore, it is important to include ingredients that boost microbial establishment and activity (ABUBACKR et al., 2014). This allows an adequate transition from a liquid to a solid forage-based diet, which in turn allows early weaning without any negative effects on daily weight gain (RASBY, 2007). The establishment of bacterial species in lactating beef calves in the tropics has been scarcely studied. Traditional techniques for cultivating microorganisms, including isolation, characterization, and cell count, have provided information; however, information on the diversity and identification of rumen bacteria are still limited, but molecular techniques have been shown to be more efficient for bacteria species identification (JAMI et al., 2013;JESUS et al., 2015). This study evaluated the effect of calf starter intake on, ruminal development, productive responses and overall bacterial diversity of dual purpose calves in the tropics.

Animal, diets, and experimental design
This experiment was conducted in a cattle production facility dedicated to the study of the dual-purpose system (DPS), located in the Tapachula municipality, in Soconusco, Chiapas, Mexico, between 14° 91´ 36´´ north latitude and 92° 32´ 55´´ west longitude with an altitude of 177 m. Precipitation from November to April oscillates between 75 and 800 mm, and from May to October period, fluctuating between 1,200 and 3,000 mm. The temperature fluctuates between 18 °C and 34.5 °C. Chemical analyses of experimental samples were conducted in the Nutrition Laboratory of the University Center of the South of University of Guadalajara, located in Ciudad Guzman, Jalisco, Mexico.
Twelve male Brahman × Swiss American cross bull calves from a dual-purpose system were randomly assigned two treatments, with six animals per treatment: milk-fed calves + Taiwan grass (Pennisetum purpureum, MT) and MT + calf starter, (MTS). The calf starter was devised according to the nutritional requirements of beef cattle (NRC, 1996). The proportion of ingredients per 100 g, as offered, was as follows: soybean meal 27 g; ground corn 48.60 g; palm kernel meal 22 g (by-product derived from palm coconut oil extraction of Elaeis guineensis); mineral premix 2 g (per 100 g containing: sodium 9.60 g, chlorine 14.40 g, calcium 21.14 g, sulfur 5.20 g, magnesium 0.80 g, zinc 0.42 g, manganese 0.26 g, cobalt 10 mg, iodine 4.64 mg, and selenium 0.40 mg); yeast culture 0.30 g (Saccharomyces cerevisiae, live culture yeast cells: 2 x 10 9 CFU g -1 ); and microminerals 0.10 g (selenium methionine 590 ppm, zinc di-lysine 3,000 ppm, iodine peptide 30 ppm, cobalt peptide 30 ppm, chromium methionine 990 ppm, copper di-lysine 500 ppm, manganese di-lysine 3,000 ppm, iron di-lysine 1,500 ppm, and vitamin E 50 IU kg -1 ). Taiwan grass was harvested every day, freshly chopped, and offered to the calves at 0700 h and 1300 h. All animals had unlimited access to clean water and chopped Taiwan grassand werehoused in individual pens, additionally, the calves in the MTS treatment had free access to the calf starter at 08:00 and 13:00 h.

Chemical analysis
The experimental feed sare shown in Table 1. Samples were dried in a circulating air oven at 60 °C for 24 h and then milled in a hammer mill equipped with a 2mm sieve for further analysis. Total dry matter (DM) was determined using a circulating air oven. Crude protein (CP) was determined by the Kjeldahl method, and ethereal extract (EE) and ash content, after incineration of the samples in a muffle at 550 °C, were calculated from the difference using the technique described by AOAC (2007).
The determination of the fiber fractions, neutral detergent fiber (NDF) and acid detergent fiber (ADF), was performed using alpha amylase without ash correction, as specified by VAN SOEST et al. (1991).

Handling and feeding
The calves suckled the mother's colostrum in the first three hours of life, the navel was disinfected with 3% iodine solution, and the calf weight was recorded with a digital scale with 300 kg capacity and 0.1 kg accuracy no later than 24 h after birth. The calves remained with the mother constantly during the first 20 days, and, from day seven, the milking of the residual milk began. From day 21 onward, after the morning milking (08:00 h), all calves were allowed to nurse the dam for 30 min and were subsequently separated from the mother until the next day. The calves were weaned at seven months of age (210 d) in the MT treatment and at 90 days of age in the MTS treatment to determine possible differences in rumen development by the effect of the calf starter. At 90 d of age, was orally administered an endoparasiticide, a single dose of 2 mL per 45 kg of live weight (each 100 mL contained 13.0-g albendazole and 1.0 g of cobalt sulfate). Five milliliters of vitamins were administered intramuscularly (each mL contained 500,000 IU of vitamin A palmitate, 75,000 IU of vitamin D3 cholecalciferol, 50 IU of vitamin E tocopherol, and 1,560 μg of phosphorus).

Growth performance and calf starter feed intake
Calves were weighed at birth at 7-day intervals until 210 d of age, weight was recorded and the average weight gain (ADG) was calculated weekly. Consumption of calf starter was measured daily throughout the experiment to estimate the intake and the remaining feed was weighed at each delivery time 24 h later.

Ruminal papillae development
Three calves were harvested from each treatment at 90 days of age, according to the Norma Oficial Mexicana promulgated in the Official Journal of the Federation (NOM-044-ZOO-1995). The digestive system, from the esophagus to the large intestine, was removed. A portion from the rumen and to the beginning of the small intestine was dissected, the contents of the compartments were emptied, washed with water until clean and free of digestive content, and samples of ruminal tissue were collected. To complete 20 observations per rumen cavity per treatment, each sac of the rumen (cranial, ventral, dorsal and dorsal blind) was dissected into sections of approximately 16 cm 2 ; four sections of each sac were taken at random and, from each of them, five samples of approximately 1 cm 2 were removed to evaluate the number of papillae per cm 2 (NP), length (LP), and papillae width (WP).

Number of ruminal papillae per cm 2
Each sample was placed on a slide and observed under a microscope with a 4x lens and a graduated ocular with an observation field of 9 mm 2 and precision of 0.025 mm. In the center of each 1 cm 2 sample, we counted the total papilla in an observation area of 9 mm 2 , and we calculated the number of papillae per cm 2 from the data obtained.

The length and width of the papillae
From each 1 cm 2 sample, a papillae in the center of the observation field was randomly chosen, and the length of the papilla from the base to the apex and the width at the base was measured.

Identification and molecular characterization of ruminal bacteria
After calves euthanized, 50 mL of ruminal liquid from the cranial, ventral, dorsal, and dorsal blind sacs were taken from rumen digesta samples and were filtered three times through sterile gauze; 30 mL were then deposited in sterile vials and immediately stored at −196 °C in liquid nitrogen for rumen 16S rRNAanalysis.DNA was extracted from 50 µL of ruminal fluid using the ZymoBIOMICS™ DNA Miniprep kit (Zymo Research Corp., Irvine, CA, USA), according to the manufacturer's instructions, and the DNA concentration was determined by spectrophotometry (STULNIG & AMBERGER, 1994). The identification of ruminal bacteria genotype, were determined by the use of amplicon generation with the Illumina platform using the forward primer 5' TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG CTC CTA CGG GAG GCA G 3' and the reverse primer 5' GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GCT TGT GCG GGC CCC CGT CAA TTC 3.' After amplification, the fragments were purified using Ampure XP and a gene library was created according to sequencing by synthesis (SBS) platform, with the Illumina MiSeq system (Illumina, Analysis Software version: 2.6.2.3), in the 2 x 300 run format, which has the capacity to generate approximately 12 Gb and guarantee 20,000 readings per sample, with the use of amplicons. Once the libraries were generated, quantification, preparation of the equimolar pool of the libraries, sequencing run, data separation for analysis and bioinformatics analysis were conducted.

Statistical analysis
The data obtained from the productive and papillae development variables were subjected to using PROC GLM of SAS (Statistical Analysis System, version 9.0). Significance was set at P< 0.05. The statistical model used was where Y ij is the response variable; µ is the general mean; T i fixed effect of i th traetment (MT or MTS); δ ij ramdom effect of calf; ε ij is the error associated with measurement taken from calf j from i th treatment.

RESULTS
The calf starter intake (kg) increased from month two when the feed was offered as free access; it increased from month 2-4 as the percentage of LW, but decreased in month 5 (Table 2). In month 2, calf starter intake was 0.42 kg calf -1 d -1 and represented 0.58 % of the LW; at the time of weaning (3 months of age), it was 0.99 kg calf -1 d -1 and represented 1.22 % of the LW. In the postweaning period, it increased to 1.56 kg calf -1 d -1 in month 4 and represented 1.64 % of the LW. The birth and live weights from ages 1 to 5 months were similar between treatments (P = formal style please fix all others> 0.05), but in months 6 and 7, LW was higher (P< 0.05) in calves receiving the MTS treatment (Table 3). The ADG was similar between treatments in months 1 to 4 (P> 0.05) and was higher in MTS calves at 5 to 7 months. The average from birth to 210 days of age (P< 0.05; Table 4) was also higher in MTS. In this study, the starter feed intake improved the ADG in the post-weaning period.
The number of ruminal papillaes per cm 2 was higher in the cranial, ventral, dorsal, and caudodorsal sacs of calves fed with calf starter (P< 0.05), this was possibly due to the stimulation of ruminal papillae proliferation in various sacs (Table 5). The ruminal papillae length (LP) was higher in the rumen ventral sac of calves fed with calf starter (P< 0.05), but there was no difference between treatments in the cranial, dorsal, and caudo-dorsal sacs (Table 6). There was no difference in the papillae width in the ventral and dorsal rumen sacs (P> 0.05), but it was higher (P< 0.05) in the cranial sac of calves fed with the MT and in the caudo-dorsal sac of the calves fed with the MTS treatment ( Table 7).
The molecular identification of the ruminal bacteria, according to bioinformatics analysis, resulted in 28 different phyla; At level phylum, Bacteroidetes was the most abundant bacteria, (average relative abundance of 41 %) followed by Firmicutes (33 %), Proteobateria (5.5 %), Actinobacteria (1.38 %), Verrucomicrobia (0.80 %) and Nitrospire (0.65 %). Fibrobacteres (0.81 %) were identified in the MT group and the Tenericutes (0.57 %) in calves with MTS (Figure 1). At the genus level, 474 genera were detected. Eight genera were predominant, but their relative abundances were different. Provotella was the most abundant genus in all treatments; however, the Butyrivibrio genus was only present in the MT group and for the calves with starter feed, we observed the bacteria Desulfonauticus autotrophicus had been not previously reported in ruminants (Table 8, Figure 1).

DISCUSSION
In this study, the starter feed intake was 0.99 kg at the time of weaning (3 months of age), Table 1 -Chemical composition of the calf starter and the Taiwan grass.

Chemical composition
Calf starter Taiwan grass The calf starter intake observed in this study was lower than that reported by CASTRO & ELIZONDO (2012), who fed Holstein calves with calf starter as meal and the intake was 1 kg d -1 at 2 months of age and 1.42 kg d -1 during the transition at 42 to 55 days of age. During the transition, from lactating to weaning, the consumption of milk affects the calf starter intake (SAEGUSA et al., 2017). HAISAN et al. (2019) observed that, in Holstein calves, the intake was 1.07 kg d -1 when the milk intake was 5 L d -1 and decreased to 0.72 kg d -1 when it was 10 L d -1 . In calves from dual-purpose systems with milk restriction, this could have positive effects during the transition. As observed in this study, the quantity of feed intake was enough to produce a larger number of papillae in the different ruminal sacs of calves fed with MTS treatment, which facilitated weaning at three months of age without any effect on the ADG and a positive effect on LW and ADG in the post-weaning period.
The live weight of calves in the months 1 to 5 was similar between both groups; whereas, in months 6 and 7, it was higher in the calves fed Table 2 -Calf starter intake by beef calves from a dual-purpose system, base as offered.  Table 3 -Live weight of beef calves from the dual-purpose system, from birth to seven months of age (210 d).  SEGURA et al.(2017) reported that weight at weaning in beef calves is associated with breed, maternal ability, milk available to the calf, and handling, where the intake of calf starter is determinant because affects the weaning weight. Therefore, results of this study suggested that calves fed starter feed after 21 days of age and weaned at three months have a weaning weight similar to that of beef calves without milk restriction. SANDOVAL et al. (2005) showed that milk consumption affects the ADG of DPS calves in the first 3 months of age, and as lactation progresses, the effect of milk decreases due to the lower milk consumption. This was observed in MT fed calves, Table 4 -Average daily gain (g calf -1 d -1 ) of beef calves from the dual-purpose system, from birth to seven months of age (210 d). Table 5 -Number of papillae per cm 2 in the cranial, ventral, dorsal, and caudodorsal sacs in the rumen of calves from a dual-purpose system at three months of age (90 d).

Month of age MT MTS SEM 2 ---------------------------------------Kg-----------------------------------
- where ADG decreased after 4 months of age; whereas, in MTS fed calves, they observed a positive response due to the increase in calf starter intake showing an ADG of 949 g calf -1 d -1 in month 5 and 993 g calf -1 d -1 at 7 months. Similarly, BETANCOURT et al. (2012) reported that in lactating calves supplemented with calf starter plus a yeast culture (Saccharomyces cerevisiae), ADG was 732 g calf -1 d -1 . CASTILLO et al. (2018) reported an ADG of 450 g calf -1 d -1 in the first month of age and 1077 g calf -1 d -1 in the third month of age, when the calf starter intake was 2.9 kg calf -1 d -1 in calves weaned at 4 months of age. Calves fed with calf starter had a higher number of papillae per cm 2 in the ventral, cranial, dorsal, and caudo-dorsal ruminal sacs; it has been suggested that lactating calves fed with fermentable carbohydraterich diet, the production of butyrate is higher, which promotes the development of the ruminal epithelium (KHAN et al., 2008;SIMEONE & BERETTA 2016). This may be due to the increase in the metabolism of butyric acid in the ruminal epithelial cells (GÓRKA et al., 2009;PENNER et al., 2011), which oxidizes to β-hydroxybutyrate and is later absorbed, increasing its plasmatic concentration (LESMEISTER & HEINRICHS, 2004). CASTRO and ELIZONDO (2012) reported an average papillae length of 1.93 mm and width of 0.98 mm from different rumen cavities of eight-weekold Holstein calves that consumed 1.0 kg d -1 of calf starter as meal. SHIN et al. (2011) observed a higher papillae length (2.24 mm vs 2.54 mm) in 42-dayold Holstein calves that received sodium butyrate in the milk replacer with a starter intake of 0.4 kg calf -1 d -1 . Thus, it is important to stimulate calfstarter intake at an early age in beef calves and to include ingredients that promote the development of ruminal Table 6 -Papillae length (mm) in the cranial, ventral, dorsal, and caudodorsal ruminal sacs of calves from a dual-purpose system at three months of age (90).
-   microorganisms. In this study, the calf starter led to a positive response in the development of ruminal papillae and ADG, probably by the inclusion of palm kernel meal in the formulation.
In pre-ruminants, is difficult to study the impact of host age alone due the con-founding effect of the diet because undergo from a whole milk based to a solid based diet in a short time for the early rumen development (MALMUTHUGE & GUAN, 2017;O´CALLAGHAN et al., 2018), but, Bacteroidetes, Fimicutes and Proteobacteriahave been reported as the most dominant phyla in the rumen (WU et al. 2012). In our study, the most predominant bacteria in the MT calves ruminal fluid were Prevotella (15 %), Ruminococcus (4 %), and Blautia (4 %), whereas, in MTS, they were Prevotella (16 %), Dysgonomonas (5%), and Bacteroides (4%); however, from day 15 to 83 of age, the populations dramatically changed in response to starter feeding: Prevotella increased to 42 % and the other genera decreased to less than 0.1 %. LI et al. (2012) observed in calves fed with a milk-replacer, an increase, from the 2 th to the 6 th week of the relative abundance of Bacteroidetes, a decrease in Firmicutes and Proteobacteria. However, REY et al. (2014) evaluated the bacterial populations established in the rumen of dairy calves, from days 5 to 12, and reported that the predominant genera were Bacteroides (21 %), Prevotella (11 %), Isobacterium (5 %), and Streptococcus (4 %), those results are different to ours.
We reported the presence of the bacteria Desulfonauticusautotrophicus sp. nov. (3 %) in the rumen of calves fed MTS a species that has not previously been reported in ruminants. This bacteria was isolated from production water of an oil field in Northern Germany near Hamburg and described by MAYILARAJ et al. (2009) as a novel, moderately thermophilic and halophilic, sulfate-reducing bacterium. The cells were Gram-negative, straight to slightly curved rods, and motile by a single polar flagellum; they grew well between 40 and 62 °C, but the optimum conditions for growth were 58 °C and a pH 7.8. Both temperature and pH differ with what was reported in this study, where the ruminal temperature was approximately 38 °C and pH 6.18, but, Desulfonauticus autotrophicus, a strictly anaerobic bacterium, which is consistent with the rumen conditions, and could be related to the inclusion of palm kernel meal in the calf starter; however, the ecological role of this bacteria in the rumen microbial ecosystem, needs to be further investigated. ABUBACKR et al. (2014), using palm kernel cake to feed goats, observed an increase in the population of total and cellulolytic bacteria, such as Ruminococcus albus and Ruminococcus flavefaciens, while, the population of methanogenic and protozoan bacteria decreased; CHANJULA et al. (2010) reported that by including more than 35 % of palm meal in the diet, the concentration of volatile fatty acids decreased in the rumen.

CONCLUSION
The consumption of calf starter by calves of the dual-purpose system in the tropics, for 21 days of age, promoted a greater number of rumen papillae. Weaning at 90 days of age does not affect the live weight or average dairy gain, but in the post-weaning period, the response to the calf starter by the variables mentioned is higher. The presence of species Desulfonauticus autotrophicus sp. nov has not been previously reported in ruminants, and we considered that its presence may be due to the inclusion of palm kernel meal; however, further research is required for the isolation, identification, and description of the bacteria activity of ruminants fed with this ingredient.

ACKNOWLEDGMENTS
The results are part of the fiscal project 2018 SIGI: 13212634416, "Production of high quality and safe beef in the tropics of Mexico.

DECLARATION OF CONFLICTS OF INTERESTS
The authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analysis, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

AUTHORS' CONTRIBUTIONS
All authors contributed equally for the conception and writing of the manuscript. All authors critically revised the manuscript and approved of the final version.