Effects of Replacing Dietary Antibiotic Supplementation with Chitosan Levels on Rumen Metabolism and Nitrogen Use in Finishing Steers Fed Forage-Free Diets

The aim of this study was to evaluate the effects of replacing antibiotics with increasing doses of chitosan (CHI) on nutrient intake and digestibility, corn grain excretion, ruminal fermentation, nitrogen metabolism, and feeding behavior in finishing steers fed forage-free diets. Five ruminally cannulated crossbred steers were assigned to a 5 x 5 Latin square experimental design and given the following diets: C0= basal diet with no additives, ANT= basal diet with inclusion of virginiamycin 30 mg/kg DM, C375= basal diet with inclusion of chitosan 375 mg/kg DM, C750= basal diet with inclusion of chitosan 750 mg/kg DM, and C1500= basal diet with inclusion of chitosan 1500 mg/ kg DM. Supplementation with CHI did not affect ruminal pH and ammonia nitrogen (N-NH 3 ) concentration. Chitosan quadratically affected corn grain excretion and molar proportion of ruminal butyrate, with greater values for C750. Animals fed C375 showed greater dry matter intake and neutral detergent fiber intake; and greater digestibility of DM, starch, and NDF. Animals fed C750 and C1500 diets presented greater absorbed N compared to ANT. Animals fed ANT, C750, and C1500 spent more time eating and chewing than animals fed CON. More specifically, the dose of 375 mg/ kg DM of diet was very promising for steers fed free-forage diet due to its ability to improve nutrient digestibility, with only slight changes in N metabolism.


INTRODUCTION
In the last decade, there has been a reduction of forage concentration in finishing diets in Brazilian feedlots.In 2009, the average of using forage on a total dry matter basis was 28.8% and it dropped to 16.7% in 2019 (Silvestre & Millen, 2021).The benefits of increasing concentrate levels in cattle finishing diets and the use of forage-free diets are not only related to the increase in dietary net energy for gain, improving feed efficiency, reduction of the cost per unit of metabolizable energy, and facilitating the feedlot management by the easiness of diet handling but also because of the demand for property space for crops, storing and distributing roughage in confinement (Rivaroli et al., 2016).
The concept of a forage-free diet has been applied to diets without neutral detergent fiber (NDF) from forage.Still, when supplied to ruminants, the nutritionist must take into consideration several key points, such as i) the level of starch and type of corn grain processing, ii) the use of non-forage NDF, iii) adaptation of animals to the diet, iv) the diet particle size, v) the use of some additives to prevent ruminal metabolic problems, and vi) the precise control of animal feed intake.Otherwise, as explained by Paula et al. (2019), the rumen epithelium can be susceptible to injury, resulting in translocation of endotoxin and bacteria into the bloodstream (Emmanuel et al., 2008), with a further cascade of non-specific immune responses, infection, and abscess formation in the liver and the foot, bloat, and others disorders correlated with greater ruminal production of lactate and lower pH (Ametaj et al., 2009;González et al., 2012;Owens et al., 1998).All these alterations are associated with a decrease in animal performance and damage to animal welfare and health, impacting the system's profitability (Paula et al., 2019).
Antibiotics have been used to manipulate ruminal fermentation, preventing those problems and increasing animal performance (Tiseo et al., 2020).As a result, worries about the potential harm these compounds might do to human health and the spread of bacterial multidrug resistance have grown (Cheng et al., 2014).
Because of this, resistant bacteria are now a major threat to global health, and one of the biggest problems facing scientists today is halting their spread (Ferri et al., 2017;Van Boeckel et al., 2015).
In this sense, chitosan (CHI), a natural biopolymer extracted from exoskeletons of crustaceans and insects, has been extensively investigated.Senel & McClure (2004) explained that chitosan exhibits various biochemical properties resulting in numerous applications in diverse fields of knowledge, such as pharmaceutical, medical, agriculture, cosmetics, nutritional enhancement, textile, and food processing in livestock industry and veterinary medicine.In animal production, testing CHI in animal diets has gained enormous interest due to its properties as a natural feed additive to replace traditional antibiotics and improve growth, performance, and meat production because of its antimicrobial, anti-inflammatory, antioxidant, and digestive modulatory activity on ruminal metabolism (Dias et al., 2017;Goiri et al., 2010;Harahap et al., 2022;Jiménez-Ocampo et al., 2019).Ruminant nutritionists are very interested in CHI because it has a mode of action similar to ionophores as a ruminal modulator, altering ruminal fermentation to a more energy-efficient pathway (Seankamsorn et al., 2021(Seankamsorn et al., , 2020) ) by inhibiting Gram-positive bacteria growth and decreasing acidosis events (Goiri et al., 2010).
Several authors have reported consistent results, including a significant increase in ruminal propionate concentration and consequently improvement in DM, CP, and NDF digestibility when beef steers and dairy cows received CHI in the diet (Araújo et al., 2015;Dias et al., 2017;Vendramini et al., 2016), but none of those studies had included CHI in the ruminant forage-free diet.Thus, this study hypothesized that animal-fed forage-free diets have beneficial changes in ruminal pattern fermentation by receiving CHI in the diet and that greater levels of CHI promote better results.The technological innovation in this study is based on using an active organic ruminal fermentation modulator in diets without forage, highlighting the great metabolic challenge for animals without antibiotics to guarantee favorable conditions for ruminal fermentation.Therefore, this study aimed to evaluate the effects of replacing ANT with CHI, ranging from 0 to 1,500 mg/kg on nutrient intake and digestibility, ruminal fermentation, and nitrogen use of finishing steers fed forage-free diets.

MATERIALS AND METHODS
All animal care procedures adopted in this study were conducted following the Institutional Animal Care and Use Committee Guidelines of the Federal University of Grande Dourados (approval protocol: 023/2015 CEUA/UFGD).

Experimental Site, Animals, and Diet
This study was conducted at the Ruminant Nutrition facility and Animal Nutrition Laboratory, School of Agrarian Sciences of the Federal University of Grande Dourados, Dourados, Brazil.Five crossbred steers, fitted with rumen cannulas, 2 years of age (350 ± 20 kg body weight) were randomly distributed to a 5 x 5 Latin square experimental design.Prior to the beginning of this study, animals received a step-up adaptation diet for 30 days with a progressive inclusion of the concentrate in the diet, 50%, 60%, and 80%, until they reached the level of 100% concentrate based on corn grain.Later, animals were housed in individual concreted pens (8 m 2 , 2 x 4 m) with individual feed bunks of 0.5 linear m for 19 days during each experimental period (10 days for adaptation and 9 days for data collection), totaling 95 days.Animals were randomly assigned to the following treatments: C0= basal diet with no additives, ANT= basal diet with inclusion of virginiamycin 30 mg/kg DM, C375= basal diet with inclusion of chitosan 375 mg/kg DM, C750= basal diet with inclusion of chitosan 750 mg/kg DM, and C1500= basal diet with inclusion of chitosan 1500 mg/ kg DM.Chitosan doses were selected for their similarity to those already tested by our research group with animals fed on pasture (Dias et al., 2017).The chitosan used in the present study has the following technical specifications: ≥ 850 g/kg deacetylation degree, 0.32 g/mL density, pH 7.90, viscosity < 200 cPs, total ash 1.35 g/100 g, and loss on drying 9.3 g/100 g (Polymar Indústria e Cia.Imp. and Exp.LTDA, Fortaleza, Brazil).
Animals were fed ad libitum twice a day at 8 AM and 3 PM with a diet containing 85% corn grain and 15% pellet composed of protein, minerals, and vitamins (Table 1), with no forage inclusion.Ingredients and orts of each diet were weighed every day to estimate individual dry matter intake (DMI); feed was adjusted daily to allow for a minimum of 5%-10% orts.

Sample Collection and Chemical Analysis
Dietary ingredients and orts were sampled weekly, pooled according to the period and animal, and stored at -20 °C for chemical composition and nutrient intake estimation.Fecal samples were collected from day 11 to day 15 of each experimental period directly from the rectum at 8 AM, 10 AM, 12 PM, 2 PM, and 4 PM.Ingredients, orts, and feces samples were dried in a forced air oven at 65 °C for 72 h, ground to pass through a 1-mm screen, and then pooled according to the fecal excretion, chemical composition, and nutrient digestibility period.Diet ingredients, orts, and feces samples were analyzed according to AOAC (Official Methods of Analysis, 2020) for DM (method 930.15), crude protein (N x 6.25; method 984.13), and ash (method 942.05).The neutral detergent fiber was analyzed using α-amylase without sodium sulfite (Mertens, 2002).Acid detergent fiber was analyzed according to Van Soest et al. (1991).Starch concentration was analyzed following the methodology proposed by Gandra et al. (2016).Fecal excretion was estimated using titanium dioxide (TiO 2 ) as an external marker added daily to the diet between day 5 and day 15 of each experimental period, being the first five days for adaptation of the animals to the TiO 2 on diet and the last five for fecal collection, following the methodology described by Dias et al. (2017).Fecal TiO 2 was analyzed by UV/Vis spectrophotometry using the methodology previously described by Myers et al. (2004).Total DM fecal excretion was calculated as follows: FE = TIS ÷ TICF where FE is the daily DM fecal excretion (g/d), TIS is the titanium dioxide supplied (g/d), and TICF is the titanium dioxide content in feces (g/g DM).

Corn Grain Excretion
Feces samples were collected directly from the animal's rectum on days 15, 16, and 17 of each experimental period, before the morning and afternoon meals, to evaluate the corn grain excretion.Approximately 400 g of feces were weighed, and then samples were washed on a 4-mm sieve as described by Rennó et al. (2015).Corn grain particles were manually collected and weighed to evaluate the corn grain excretion.

Nitrogen Balance and Microbial Protein Synthesis
Urine samples (100 mL) were collected by preputial stimulation on day 16 of each experimental period, four hours after the concentrate supply.Urine aliquots (10 mL) were diluted in sulfuric acid (40 mL; 0.036 N H 2 SO 4 ) to prevent purine derivative degradation and uric acid precipitation; then, diluted urine samples were analyzed for creatinine, urea, uric acid, and allantoin.Allantoin was assessed according to Vendamini et al. (2016), creatinine and uric acid were determined by colorimetric methods using commercial kits (Labtest, Lagoa Santa, Brazil; Gold Analisa Diagnostica Ltda, Belo Horizonte, Brazil), and readings were taken using a semi-automatic biochemical analyzer Bioplus,Barueri,SP,Brazil).Daily urinary excretion was estimated as follows: Urinary excretion= (27.36 * BW) ÷ creatinine concentration in urine (Paiva et al., 2016).
The total excretion of purine derivatives (PD) was calculated as the sum of allantoin and uric acid excreted in urine.The absorbed microbial purines (Pabs, mmol/ day) were calculated using the equation: Pabs = (PD-0.236*BW 0.75 ) / 0.84 where 0.84 is the recovery of purines absorbed as purine derivatives and 0.236*BW 0.75 is the endogenous excretion of purine derivatives (Paiva et al., 2016).
Ruminal synthesis of nitrogen compounds (Nmic, gN/day) was calculated based on the absorbed purines (Pabs, mmol/day) using the equation (Chen & Gomes, 1992): where 70 is the N content in purines (mgN/mol), 0.134 is the purine N: total N ratio in bacteria (Paiva et al., 2016), and 0.83 is the intestinal digestibility of microbial purines.

Urea and Creatinine Metabolism
On day 16 of each experimental period, blood samples (20 mL) were collected by puncturing coccygeal vessels four hours after the diet supply.Then, samples were centrifuged at 2,700 x g for 20 min (4 °C), serum supernatant was harvested and analyzed for urea and creatinine by the colorimetric method using commercial kits (Gold Analisa Diagnostica Ltda, Belo Horizonte, MG, Brazil) and readings were taken by a semi-automatic biochemical analyzer Bioplus,Barueri,SP,Brazil).Plasma depuration or clearance of creatinine and urea was obtained by the ratio of the urinary excretion for 24 hours to the plasmatic concentration of each substance.Urea nitrogen and creatinine nitrogen excretion were analyzed by their total urinary concentration multiplied by 0.466 or 0.3715, respectively.

Ruminal Fermentation
Samples of ruminal fluid (around 50 mL) were collected through the ruminal cannula and filtered on day 17 of each period before the morning feed and 2 h, 4 h, 6 h, and 8 h after the morning meal.Immediately after collection, the ruminal fluid pH was determined using a potentiometer (pH 1500, Instrutherm, São Paulo, SP, Brazil).Short-chain fatty acid (SCFA) were determined by mixing 2 mL ruminal fluid aliquots, collected from all animals, with methanolic acid (400 μL; 98%-100% H 2 CO 2 ), and then centrifuging at 7.000 x g at 4 °C for 15 min.The supernatant was collected and stored at −20 °C for posterior analysis of ruminal SCFA by gas chromatography (Shimadzu GC 2010 with automatic injection) as previously explained by Dias et al. (2017).Briefly, the equipment was equipped with a split injector, dual flame ionization detector (temperature at 250 °C), and capillary column (30 m length and 0.53 mm internal diameter; Stabilwax, Restek, Bellefonte, PA, USA) at 145 °C.The start temperature was 40 °C followed by 40 °C/min increments until 120 °C was reached.In sequence, the temperature increments were 10 °C/min until a temperature of 180 °C was reached, and from this temperature to 240 °C, the increment was 120 °C/min, which was maintained for 3 min.Two standards were used to identify fatty acids: WSFA-2 (Ref.47056, Supelco, Bellefonte, PA, USA) and glacial acetic acid (Ref.33209, Sigma-Aldrich©).SCFA concentration was calculated using the GC solution v. 2.42.00 software (Shimadzu©).
For ammonia nitrogen (NH 3 -N) concentration measurements, 2 mL ruminal fluid was mixed with sulfuric acid (0.5 mol/L H 2 SO 4 ) and stored at −20 °C for posterior analysis following the method of Paiva et al. (2016).

Feeding Behavior
The feeding behavior of animals was evaluated on day 18 of each experimental period by visual observations at 5-min intervals of the variables: time spent on feeding (h/d), ruminating (h/d), feed intake frequency, water intake, and idleness (Johnson & Combs, 1991); beginning at 8 AM, during 24 hours as previously described by Ferrari et al. (2019).For better data collection, steers were adapted to artificial light at night for 3 days before the evaluation.

Statistical Analysis
Statistical analysis tests for normality and homogeneity were performed using PROC UNIVARIATE in SAS 9.2 software (SAS Institute Inc., Cary, NC, USA, 2009).Differences between ANT and all other treatments individually (CON, C375, C750, and C1500) were tested by the Dunnett test, and differences between CHI treatments (CON, C375, C750, and C1500) were tested by polynomial regression using the PROC MIXED procedure.Results are presented as LSMEANS and significance was declared p≤0.05.Data were analyzed according to the model: where, Yijk is the dependent variable, μ is the overall mean, Di is the fixed effect of diet, P j is the random effect of the experimental period, A l is the random effect of the animal, and e ij is the residual error.
Ruminal fermentation data were analyzed as repeated measures using the REPEATED from PROC MIXED following the model below: where, Y ijk is the dependent variable; μ is the overall mean, D i is the fixed effect of diet, t j is the fixed effect of time of measurement, D i (t i ) is interaction, and e ij is residual error.N stands for Gaussian distribution, s2v is the variance associated with experimental diets, MVN stands for multivariate normal, and R is the variance-covariance matrices of residuals due to the repeated measurements.It was evaluated the following variance-covariance matrix (CS, CSH, AR(1), ARH(1), TOEP, TOEPH, UN, FA(1), and ANTE(1)).The matrix was chosen using a Bayesian method.Experimental period and animals were used as random effects.

RESULTS
There was no linear or quadratic effect (p>0.05) according to CHI levels included in the diet on nutrient intake and digestibility, and fecal pH (Table 2).However, a quadratic effect (p= 0.037) was found for corn grain excretion, with the highest values for animals fed C750 (Table 2).When CHI treatments were compared to ANT, the C375 diet showed greater DM and NDF intakes (p<0.05),480 g/day and 90 g/day, respectively.
Regarding nutrient digestibility, animals fed C375 had greater DM, starch, and NDF digestibility compared to the ANT, while the treatments C750 and C1500 presented similar intake and digestibility to the ANT diet.Fecal pHs were greater (p<0.05) for C0, C375, and C750 diets compared to the ANT.Animals fed ANT and C1500 had lesser (p<0.05)fecal pHs compared to the other treatments (Table 2).
Analyzing ruminal fermentation, CHI did not affect ruminal pH, NH 3 -N, and total SCFA molar proportions, but CHI linearly affected (p= 0.038) the molar proportion of ruminal acetate and quadratically affected butyrate (p= 0.005), being the highest values found for the C750 diet (Table 3).In addition, the molar proportion of isovalerate was significant with a quadratic effect (p=0.002), with the highest value found for the ANT group (Table 3).
No differences existed in the molar proportion of ruminal propionate, butyrate, and isobutyrate when CHI treatments were compared to animals fed ANT.Nonetheless, compared to animals fed ANT, the molar proportions of ruminal acetate were lower for animals fed C0 and C375 (p<0.05), while molar proportions of ruminal valerate were lower for animals fed C0 and C1500 (p<0.05).The molar proportions of isovalerate presented lower values for the treatments C375 and C750 (Table 3).
Neither linear nor quadratic effects (p>0.05) were detected for nitrogen (N) intake, excretion, balance, or microbial protein synthesis (Table 4).However, compared to the ANT diet, animals fed C0 diet had greater (p<0.05)fecal N excretion and lower absorbable N, while animals fed C375 presented a greater (p<0.05)urine N excretion, resulting in lower retained N (p<0.05),but no changes on microbial N and protein synthesis compared to the ANT (Table 4).On the contrary, C750 diet decreased (p<0.05)urine N excretion and consequently improved the N retained compared to animals fed ANT.Despite these changes in N excretion and balance among treatments, microbial protein synthesis parameters were only greater for C1500, compared to ANT (Table 4).However, the other CHI inclusion levels presented similar results compared to the ANT group (Table 4).
Animals fed C375 presented greater (p<0.05)values of serum urea, creatinine, N-urea, and N-creatinine, lower urea excretion, clearance of urea and creatinine, and lower urea excretion, compared to the ANT.Additionally, compared to the ANT, animals fed C750 had lower (p<0.05)urine concentrations of urea and urea-N, while the C1500 presented lower (p<0.05)urine concentrations of creatinine and N-creatinine but similar responses to the serum concentrations were observed between C750 and C1500 vs ANT, where creatinine and creatinine-N were higher (p<0.05)compared to the ANT, with lower (p<0.05)urea excretion and clearance (Table 5).
Chitosan quadratically affected water intake (p=0.039), with a lower value for animals fed C750, and frequency of visits to the feeder (p=0.027), with a greater value for animals fed ANT (Table 6).Compared to the ANT, animals fed C375 spent less time (p<0.05)feeding, chewing, and lower frequency of visits to the feeder, but longer time ruminating.Animals fed C750 and C1500 diets spent a longer (p<0.05)time ruminating and a lower (p<0.05)frequency of visits to feeder compared to the ANT (Table 6).

DISCUSSION
To the best of our knowledge, no data are available regarding the effect of increasing levels of CHI and comparing the animals fed antibiotics to finishing steers fed forage-free diets on nutrient intake, ruminal digestibility, fermentation patterns, and nitrogen metabolism.
Forage-free diets can decrease feed intake, resulting from the decrease of saliva flow and ruminal motility (Nagaraja & Titgemeyer, 2007), leading to some metabolic disorders.However, in the present study, animals showed similar DMI compared to the other studies that also fed beef cattle forage-free diets (Contadini et al., 2017;Paula et al., 2019).
The fact that steers were fed whole corn grain, which has a slower rate and less ruminal starch digestion than processed maize, may indicate that they should have acclimated to diets devoid of forage.Furthermore, according to Owens & Soderlund (2006), the pellet can offer some non-forage NDF that stimulates ruminal motility and rumination to some extent, avoiding a significant ruminal pH decrease.
In general, our data showed that steers fed 375 mg/kg DM presented greater DM, starch, and apparent total digestibility of NDF compared to animals fed ANT, which might be due to the greater intakes of DM and NDF for the C375 diet since up to some point the DMI drives the digestibility.This improvement in nutrient use by animals fed the C375 diet may lead to benefits in animal performance.However, even presenting a greater starch digestibility, animals fed the C375 diet had no improvement in ruminal propionate concentration and similar results regarding ruminal fermentation, urea metabolism, and nitrogen balance.
Several lines of evidence suggest the beneficial effects of CHI-supplemented diet on nutrient digestibility in cattle.In general, some authors suggested that the increased digestibility provided by chitosan addition to ruminant diets is related to its positive effect on ruminal microbiota and fermentation processes since chitosan has a strong antimicrobial effect against Gram-positive rather than Gram-negative bacteria, avoiding pH drops and decreasing ruminal protein deamination, prevent-ing nitrogen losses, but without effect on protein totaltract digestibility (Araújo et al., 2015;Dias et al., 2017;Mingoti et al., 2016;Pereira et al., 2018).Dias et al. (2017) also evaluated the increasing levels of chitosan but feeding grazing beef and reported greater DM, CP, and NDF intakes in steers fed intermediate levels (400 mg/kg and 800 mg/kg of DM) and observed a linear increase in DM and CP digestibility.Gandra et al. (2016) supplemented confined heifers and reported an increase in DM digestibility and decrease in DM and NDF intakes.In turn, Araújo et al. (2015) studied steers fed CHI and reported an increase in DM digestibility, which was unrelated to changes in feed intake.All the changes in ruminal microbiota can lead to the described changes in nutrient apparent digestibility due to CHI supplementation, which further results in beneficial changes in ruminal pH, ammonia N, SCFA production, and improves microbial protein synthesis.Especially because the basal diet used in the present study (forage-free) could lead to greater lactate production in the ruminal, pH drop, inflammation of ruminal papillae, and consequently, imbalance in ruminal fermentation, and impair the N balance and microbial protein synthesis.Nevertheless, those variables were not affected, showing that animals fed ANT or CHI (regardless of the dose) presented similar ruminal metabolism, suggesting that ANT can be replaced with CHI in the forage-free diet for steers.
According to the literature, responses to CHI supplementation can range widely from no effect on ruminal metabolism (Kirwan et al., 2021;Vendramini et al., 2016) to important changes in ruminal fermentation parameters (Araújo et al., 2015;Del Valle et al., 2017;Dias et al., 2017;Goiri et al., 2010).Araújo et al. (2015) fed Nellore steers a 60:40 roughage:concentrate ratio, and Dias et al. (2017) supplemented cattle on pasture system, and found a linear increase in molar proportion of ruminal propionate by increasing doses of CHI.Paiva et al. (2016) fed Holstein cows in lactation with increasing doses of CHI and reported a linear increase in ruminal propionate and consequently a decrease in the acetate:propionate ratio.The same authors also found a linear reduction in the ruminal isobutyrate and isovalerate by CHI inclusion, which indicates a reduction in ruminal amino acid deamination (Paiva et al., 2016).
Those differences between studies are probably related to diet composition, which plays a key role in ruminal fermentation and metabolism.Despite no changes in the ruminal NH 3 -N production, slight changes in N metabolism were observed between CHI and ANT diets.In general, animals that were given C375 diet presented greater urine N excretion, and lower retained N, showing less efficiency in regulating N balance when compared to animals fed ANT.Nevertheless, no changes in microbial N and protein synthesis were observed in these two treatments, showing that N metabolism did not disadvantage ruminal fermentation patterns and protein synthesis.Kirwan et al. (2021) explained that when consumed N is not used by ruminal microorganisms for protein synthesis, it is degraded to NH 3 -N, metabolized to urea in the liver and excreted via urine, which is an economic loss to the system with environmental impacts.The same authors explained that the lower use of dietary CP by CHI supplementation might be due to the defaunation of protozoa caused by its antimicrobial properties since protozoa play an important role in ruminal protein metabolism.However, the present study found no difference in CP digestibility or ruminal NH3-N concentration among treatments.Although this process is not fully understood, it is important to highlight that the diet used in this study does not include forage, which might be responsible for some differences in ruminal metabolism compared to the other studies.
Del Valle et al. (2017), Araújo et al. (2015), and Mingoti et al. (2016) also did not detect an influence of CHI levels on microbial protein synthesis.N metabolism data are still inconsistent in the literature regarding increasing CHI supplementation.Pereira et al. (2018) added increasing levels of CHI to the lamb diet and reported that the use of 272 mg chitosan/kg BW showed to be less efficient in regulating N balance, damaging microbial protein synthesis compared to the lower level of CHI (136 mg chitosan/kg BW).The authors explained that with higher levels, CHI tend to impair the development of not only Gram-positive, but also Gram-negative bacteria specialized in ruminal ammonia production, leading to defaunation, and decreasing N balance and microbial protein synthesis.Dias et al. (2017) supplied increasing levels of CHI to beef cattle and reported a quadratic effect on the absorbed purines and microbial nitrogen production.The authors explained that CHI antimicrobial properties may impair ruminal fermentation at a level of 1,600 mg/ kg DM concentrate.Nevertheless, in the present study, the highest level of CHI increased the microbial N and protein synthesis compared to animals fed the ANT diet, but with no difference from the other diets with CHI inclusion at different levels.Those inconsistencies among studies might be due to the differences in the basal diet composition, responsible for several changes in ruminal metabolism, and as previously commented, no other study had evaluated CHI supplementation in beef cattle fed a forage-free diets.
The excretion (mg/kg body weight) and urea and creatinine clearance were quadratically affected, with lower values in the C375 and C750 diets.These changes are generally attributed to changes in N intake or CP digestibility (Garcia-Rodriguez et al., 2015) since urea excretion in kidneys is influenced by its concentration in blood, which in turn is influenced by animal dietary conditions.However, in the present study, none of these variables were affected by CHI increasing levels.Dias et al. (2017) also observed that increasing levels of CHI lead to a quadratic effect on urea clearance and fractional excretion.
Not surprisingly, nitrogen and creatinine metabolisms were different between animals fed ANT and CON since animals fed a diet with no additive presented greater fecal N excretion and greater urinary concentrations of urea and urea-N, which can be explained by lower CP digestibility for animals fed C0 and consequently, presenting a lower ruminal efficiency on N metabolism, resulting in N loss through feces and later through urine.Araújo et al. (2015) explained that ANT and CHI treatments are expected to decrease N excretion via feces through greater CP digestibility and better N use.
Feeding behavior is extremely important when animals are given a forage-free diet since different feed additives might not stimulate chewing and rumination to the same extent, which in turn are correlated with ruminal pH and fermentation patterns, keeping animals healthy.In the present study, animals fed the ANT diet presented longer rumination time and higher frequency of visits to the feeder, which could suggest better patterns of ruminal fermentation and motility; however, those changes did not affect ruminal health, as indicated by ruminal pH and SCFA concentration.Despite improving salivation and keeping high ruminal pH, ruminating is important to control the utilization of roughage and limit the use of low-quality feed ingredients, which comprise the productive performance of animals (Huzzey et al., 2013).However, in the present study, no forage was included in the diet, so the corn pericarp and non-forage NDF of the pellet might have stimulated salivation and rumination, keeping the rumen healthy (Paula et al., 2019).Haraki et al. (2018) reported that heifers fed CHI had lower feeding efficiency of DM and NDF and chewing efficiency of NDF; the authors explained that it might be due to the lower DM intake combined with the shortest time spent ruminating standing for the CHI treatment compared to the other treatments.

CONCLUSION
In general, this study provides evidence that CHI can replace ANT in a forage-free cattle diet since, despite different responses among these treatments, CHI provides a positive result in rumen and N metabolism.More specifically, the dose of 375 mg/kg DM of diet was very promising for steers fed free-forage diet due to its ability to improve nutrient digestibility, with only slight changes in N metabolism.

Table 2 .
Nutrient intake, total tract digestion, fecal pH, and corn grain output of steers fed forage-free diets with increasing levels of chitosan

Table 3 .
Ruminal pH, N-NH₃, and short chain fatty acids molar proportions of steers fed forage-free diets with increasing levels of chitosan

Table 4 .
Nitrogen intake, output, balance, and rumen microbial protein synthesis of steers fed forage-free diets with increasing levels of chitosan Note: ANT= basal diet with the inclusion of virginiamycin 30 mg/kg DM; C0= basal diet with no additives; C375= basal diet with the inclusion of chitosan 375 mg/kg DM; C750= basal diet with the inclusion of chitosan 750 mg/kg DM; and C1500= basal diet with the inclusion of chitosan 1500 mg/kg DM. ¹Linear (L) and quadratic (Q) effect.Dunnet test: means in the same row with different superscript differ significantly (p<0.05).

Table 5 .
Urea and creatinine metabolism, output, and clearance of steers fed forage-free diets with increasing levels of chitosan Dunnet test: means in the same row with different superscript differ significantly (p<0.05).

Table 6 .
Feeding behavior and frequency of steers fed forage-free diets with increasing levels of chitosan