Palatability and digestibility of horse diets containing increasing levels of citrus pulp

Objective. To evaluate the impact of citrus pulp on the palatability and digestibility of horse diets and the physicochemical characteristics of the feces. Materials and methods. The diets were composed of 60% of the energy from the forage and 40% from the concentrate, with increasing levels of citrus pulp. Two different tests were performed. The first test assessed the palatability of concentrates. Using a randomized experimental design, 15 horses were observed for 10 days. The variables recorded were first action, first choice and intake ratio (IR). Five horses were arranged in 5x5 Latin Square design for the test assessing digestibility and fecal physicochemical characteristics. The apparent digestibility of the nutrients and the color, consistency, pH and buffering capacity (BC) of the feces were evaluated. Results. The addition of increasing levels of citrus pulp had an effect (p≤0.001) on first action. A difference was also observed in first choice, and the addition of 0, 7 or 14% of citrus pulp was preferred. A difference between treatments (p≤0.001) was also observed for IR, and the control concentrate was consumed the most. The amount of citrus pulp included had no effect (p>0.05) on the digestibility of nutrients, fecal consistency and color, and there was no effect (p>0.05) on fecal pH and BC. Conclusions. Horses can identify the presence of citrus pulp in concentrates but prefer concentrates without added citrus pulp. Citrus pulp does not negatively affect the digestibility of concentrates or the physicochemical characteristics of the feces; thus, citrus pulp is a viable alternative ingredient in the formulation of horse diets.


INTRODUCTION
The horses can use large quantities of forage to meet their nutritional requirements (1).However, diets with a high percentage of grains are used to maximize productivity (2).Currently, researchers are searching for alternative ingredients that can replace traditional ingredients such as corn, soybean meal and wheat bran for use the alternative ingredients, easily fermentable fibers, or "super fibers", such as citrus pulp, seem especially promising (3).However, only a few studies have been published regarding the use of citrus pulp in horse feed (4-7).Ott et al (4) investigated the acceptability and digestibility of diets with increasing levels of citrus pulp (0, 15 and 30%) and observed that diets with 30% citrus pulp were rejected by horses.Nevertheless no difference was observed in the digestibility of the nutrients.Manzano et al (5) observed an increase in the digestibility of dry matter and crude protein with the addition of citrus pulp; these authors using no more than 15% citrus pulp to prevent a decrease in palatability.
Another important factor to determine the digestibility of horse diets is the physicochemical characteristics of the feces, as: Feces parameters (8) such as color (9), consistency (10), pH (8) and buffering capacity (BC) (11) are important to evaluate the gastrointestinal tract health (10).
The aim of this study was to evaluate the effect of increasing levels of citrus pulp in concentrate on the palatability and apparent digestibility of horse diets, as well as the physicochemical characteristics of the feces.
El objetivo de este estudio fue el de evaluar el efecto de niveles cada vez mayores de pulpa de cítricos en el concentrado sobre la digestibilidad y palatabilidad de los alimentos para caballos, como también sobre las características fisicoquímicas de las heces.
Moreira -Palatability and digestibility of citrus pulp for horses

MATERIALS AND METHODS
Study site.This study was performed in the Equine Production section of the Pirassununga Administrative Campus of Animal Science and Food Engineering School, University of São Paulo.
Two experiments were conducted: a palatability test (test 1) and a test of apparent digestibility and physicochemical characteristics of feces (test 2).
Animals and experimental design.Fifteen horses, 10 years old, and 400 kg body weight, were used in the study.The adaptation period lasted five days and the experimental period lasted 10 days; protocol was performed once a day totalizing 150 observations.The animals were observed by instantaneous focal sampling for three consecutive minutes, and an ethogram to organize the dates.It was tested concentrates with 0, 7, 14, 21 or 28% citrus pulp (Table 1).

MATERIALES Y MÉTODOS
Lugar del estudio.Este estudio se realizó en la sección de Producción Equina del Campus Administrativo Pirassunga de la Facultad de Ciencias Animales e Ingeniería de Alimentos, Universidad de São Paulo.
Un modelo mixto lineal con esquema factorial de 5x10 se utilizó para analizar la TI, teniendo en Before starting the trial, the animals received 1 kg of concentrate (different from concentrates to be tested) in stall one hour before the test and were then led individually to the arena to conduce the trial.The troughs containing 200 g of each concentrate were arranged in five equidistant points, and the treatments and animals were randomized each day.
The preference test was used to assess the palatability of the concentrates (12).The first action (1 -sniffed, 2 -ate), first choice (first concentrate actually consumed) and intake ratio (amount of each concentrate ingested) were evaluated.
The leftovers were removed and weighed, and the intake ratio (IR) was calculated using the equation below: IR = concentrate (% citrus pulp) / (intake of 0% concentrate + intake of 7% concentrate + intake of 14% concentrate + intake of 21% concentrate + intake of 28% concentrate).
Statistical analysis.The experimental design was a split pot.A chi-square goodness of fit test was used to evaluate the first choice and first action, under the null hypothesis that all treatments would occur with equal probability.
A linear mixed model with a 5x10 factorial scheme was used to analyze the IR, taking into consideration the fixed effects of the treatment and the day, as well as the random effect of the animal, in addition to the effect of the residue.
In the case of a significant interaction, Student's t test was used to assess each treatment within each evaluation day.All analyses were performed using the Statistical Analysis System software, version 9.1.3.(13).
Animals and experimental design.Were used five nonpregnant mares, Brasileiro de Hipismo breed, weighing 406.6±76.86kg and aging six years were housed in individual stalls and arranged in 5x5 Latin Square design.The experimental period included seven days of adaptation followed by three days of data collection and analysis.
The experimental diets were formulated to meet the maintenance requirements for horses, as recommended by the National Research Council, 2007 (14).The diet were composed by 60% of the energy coming of forage (Coast Cross hay) and 40% came from concentrate with increasing levels of citrus pulp (0, 7, 14, 21 and 28%), (Table 1 and Table 2).
To evaluate the physicochemical characteristics of the feces, the samples were collected for three days in plastic bags.Subsequently, the pH was measured using a bench pH meter, and the BC was measured as described in Zeyner et al (11).
The consistency of the feces was evaluated as described by Berg et al (9), and the feces color was evaluated as described by Godoi et al (8).
Laboratory methods.A composite sample from each animal was obtained at the end of the experiment and sent to the Bromatology laboratory at the FZEA/USP.The dry matter (DM), mineral matter (MM), ether extract (EE) and cuenta los efectos fijos del tratamiento y del día, además del efecto aleatorio del animal y el efecto del residuo.En caso de haber una interacción significativa, se utilizó la prueba t de Student para evaluar cada tratamiento dentro de cada día de evaluación.Todos los análisis de llevaron a cabo utilizando el software Statistical Analysis System, versión 9.1.3(13).
Moreira -Palatability and digestibility of citrus pulp for horses crude protein (CP) contents of the concentrates, hay and feces were determined according to the procedures described by the Association of Official Analytical Chemistry (15).The acid detergent fiber (ADF), neutral detergent fiber (NDF), cellulose (CEL) and hemicellulose (HEM) contents were analyzed as described by Van Soest et al (16).The gross energy (GE) was determined using a bomb calorimeter.
We further determined the digestibility coefficients of DM, OM, CP, EE, NDF, ADF, CEL, HEM and GE.

Statistical analysis.
To evaluate the digestibility, we used the 5x5 Latin Square design using five animals fed with five diets with increasing levels of citrus pulp, repeated in five consecutive periods.The quantitative parameters were analyzed according to the statistical model: Where, Y ijk represents the dependent variable µ, the overall mean Análisis estadístico.Para evaluar la digestibilidad utilizamos el diseño Cuadrado Latino de 5x5 utilizando cinco animales alimentados con cinco dietas con niveles cada vez mayores de pulpa de cítrico, repitiéndolo en cinco periodos consecutivos.Los parámetros cuantitativos fueron analizados de acuerdo al siguiente modelo estadístico: En donde, Y ijk , representa la variable dependiente µ, la media global A i , el efecto del animal (i = 1 a 5) T j , el efecto del tratamiento (j = 1 a 5) P k , el efecto del periodo (k = 1 a 5) y ε ijk , el error residual.
Se utilizó la función PROC GLM del paquete estadístico del SAS para Windows (13) para el análisis estadístico.Las medias se compararon utilizando la prueba de Turkey con un nivel de probabilidad del 5%.A i , the effect of the animal (i = 1 to 5) T j , the effect of the treatment (j = 1 to 5) P k , the effect of the period (k = 1 to 5) and ε ijk , the residual error.
The PROC GLM function of the SAS statistical package for Windows (13) was used for statistical analysis.The means were compared by Tukey's test with a probability level of 5%.
The consistency and color of the feces were assessed using the chi-square goodness of fit test.The BC and fecal pH were assessed using the mixed linear model function of the SAS (13) program.

RESULTS
Palatability.There was a difference in first action (p≤0.001),66.91% of the animals sniffed the concentrates upon initial exposure to them, while 33.09% of the animals ate the concentrates.There was a difference (p<0.001) between treatments for the first choice variable, and treatment with 0, 7 and 14% citrus pulp were preferred (Table 3).There was a difference (p≤0.001) in IRs, as the control treatment (0%) was consumed the most, followed by treatments with 7, 14, 21 and 28% citrus pulp (Table 3).
Apparent digestibility of the diets and physicochemical characteristics of the feces.The addition of citrus pulp had no effect (p>0.05) over the digestibility coefficients of DM, CP, OM, EE, GE, CEL, HEM, NDF or ADF (Table 4).
Treatment with citrus pulp had no effect (p>0.05) on fecal consistency or color.The color of the feces remained greenish, and the consistency was classified as normal (score of 3).Adding citrus pulp had no effect (p>0.05) on BC or fecal pH.The mean values observed for BC at pH 5, BC at pH 6 and pH are shown in table 5.
Se evaluaron la consistencia y el color de las heces utilizando la prueba chi cuadrado de bondad de ajuste.La CT y el PH fecal fueron evaluados utilizando la función de modelo lineal mixto del programa SAS (13).
Las heces permanecieron de un color verdoso, y la consistencia se clasificó como normal (puntaje de 3).El añadir pulpa de cítricos no tuvo efecto (p>0.05)en la CT o en el pH fecal.Los valores medios de CT en un pH de 5, CT en un pH de 6 y de pH se relacionan en la tabla 5.  Moreira -Palatability and digestibility of citrus pulp for horses

DISCUSSION
Palatability.The first action performed by the animals upon to the initial exposure to the concentrates was sniffing (66.91%), which demonstrates that horses can select the preferred concentrate through smell (17).
This selection ability was reiterated by observation of the first choice, which demonstrated that lower levels of citrus pulp (0, 7 and 14%) were preferred (Table 3).The concentrates containing 21% and 28% citrus pulp were chosen less frequently, most likely because citrus scents are not preferred by horses (18).
In an investigation of the acceptability and digestibility of diets with increasing levels of citrus pulp (0, 15 and 30%), Ott et al (4) observed that six out of eight horses refused a diet containing 30% citrus pulp.It was not observed effect when 28% citrus pulp was added to the concentrate in this trial.
In contrast to our results, Tribucci et al (18) found no effect of different levels of citrus pulp on the first choice when studying the palatability of diets with increasing levels of citrus pulp (7, 14, 21 and 28%) in horse diets.However, they did observe an effect on the first action.The methodology used in these studies may have influenced the results, as Tribucci et al (18) used continuous troughs in a stall, while our study used separate troughs in a round pen (open area).In addition, the authors of the previous study noted a strong citrus pulp odor, which we did not observe in this study.
Assis et al (19) reported that the palatability and composition of citrus pulp can be altered by factors such as: fruit varieties, the presence of seeds, whether essential oils are removed and the processes that the fruit and fruit residues were subjected to.This variation in the composition of citrus pulp may change the taste of the final product, and could therefore
Assis et al (19) reportó que la palatabilidad y la composición de la pulpa de cítricos pueden verse alteradas por factores como: la variedad de las frutas, la presencia de semillas y si los aceites esenciales fueron eliminados en los procesos a los que se sometieron las frutas y los residuos de las frutas.Esta variación en la composición de la pulpa de cítricos puede cambiar el sabor del producto final, por lo cual podría justificar las disparidades entre los estudios sobre esta materia.
Para reforzar la clara capacidad de selección de los caballos, observamos una TI mayor en concentrados con niveles menores de pulpa de cítricos (Tabla 3) y además los animales prefirieron el concentrado de control.account for the disparities between studies of this subject.
Reinforcing the distinct selection capacity of horses, we observed a higher IR for concentrates with lower levels of citrus pulp (Table 3), and the control concentrate was preferred by animals.
Considering only the IR, the concentrate containing 0% citrus pulp was the favorite, with an IR=0.3137 (the reference value for preference was IR>0.25).This reinforces the preference for this level (0%) of citrus pulp in the concentrate.
Manzano et al (5) observed different results in their study of growing fillies when three levels of citrus pulp (0, 7.5 and 15%) were added to the diet.The diet containing 15% citrus pulp was consumed more by the fillies.The animals in our study received experimental concentrates only during the evaluation, whereas in the protocol developed by Manzano et al ( 5), the animals received hay and the test concentrates.
The use of different tests to assess palatability and acceptability greatly influences experimental results.Ott et al ( 4) and Manzano et al (5) concluded that horses had a preference for certain diets by assessing intake and leftovers, whereas Müller and Udén (20) used the first choice test.In these studies, the duration of testing and the form of assessment were the primary variables.Our study analyzed multiple factors to explain preference; thus, it is difficult to compare our results with the literature.To better understand preference and feeding behavior, it is necessary to standardize the tests used in these studies.

Apparent digestibility of the diets and physicochemical characteristics of the feces.
We observed different values for the nutrient digestibility coefficients (Table 4) compared to other studies (5-7), which may have been affected by factors such as: the forage:concentrate ratio, the composition of the diet ingredients and the quantity of food provided.
The total amount of DM provided per animal in this experiment was 2% of body weight, including at least 1% of the body weight of the animal as forage (Coast Cross hay), as recommended by the NRC (14).The forage:concentrate ratio was 60:40, based on the energy supply, representing a 75:25 DM ratio.Miraglia et al (6) analyzed the digestibility of different forage:concentrate ratios and observed that, as the percentage of concentrate increased, the apparent digestibility Considerando únicamente la TI, el concentrado con un contenido de pulpa de cítricos de 0% fue el favorito, con una TI = 0.3137 (el valor de referencia para la preferencia fue de TI>0.259).Lo anterior refuerza la preferencia por este nivel (0%) de pulpa de cítricos en el concentrado.
La cantidad total de MS suministrada a cada animal en este experimento fue el equivalente al 2% del peso corporal, incluyendo como mínimo el 1% del peso corporal del animal como forraje (heno Cross Coast), conforme a las recomendaciones del Consejo Nacional de Investigación (14).La relación forraje:concentrado fue de 60:40, con base en el suministro de energía, lo cual representa una relación de MS de 75:25.Miraglia et al (6) analizó la digestibilidad de distintas proporciones de forraje:concentrado y observó que a medida que aumentaba el porcentaje de concentrado, los coeficientes de digestibilidad de MS, materia orgánica, EB y PC también aumentaron.Cuando dichos autores utilizaron una relación de 75:25 en la dieta, al igual que lo hicimos nosotros, éstos Moreira -Palatability and digestibility of citrus pulp for horses coefficients of the DM, organic matter, GE and CP also increased.When these authors used a 75:25 diet ratio, as we did, they observed similar apparent digestibility values.The differences between the results can be attributed to different qualities of forage used in the studies.Manzano et al (5) provided food with forage:concentrate ratio of 40:60 ad libitum and observed higher nutrient digestibility coefficients than we did.
Miraglia et al (6) note that when a diet comprised of concentrate and lower quality forage favors nutrient digestibility.
The forage used in the present study had a chemical composition similar to that the Bermuda Grass hay recommended by NRC (14) and superior to the hay used by Miraglia et al (6).The hay used in our study is considered a good quality hay for horses, and accounted for a greater proportion of the total diet, which may have contributed to the consistency of the digestibility coefficients.
The amount of citrus pulp added to the concentrate did not affect the digestibility of the fiber fraction.These results are consistent with a study by Oliveira et al (2), in which pectin was added to horse diets.According to the authors, the rapid rate of food passage in horses may not have allowed sufficient microorganism activity to degrade the fiber, which resulted in no change in the digestibility of NDF and ADF.A similar observation was made by Weyenberg et al (21), who noted that the higher forage:concentrate ratio in the diet, the faster the rate of passage through the entire gastrointestinal tract.
According to Miraglia et al (6), the digestibility of the fiber fraction decreases with the addition of concentrate to the diet.We observed no decrease in fiber digestibility in this study, most likely due to the low percentage of concentrate in the diet and the quality of the hay, which provided a substrate for microorganism activity in the large intestine of horses.
We observed a GE digestibility coefficient of 46.15%, which is close to that found in diets composed solely of forage (7).This coefficient is lower than that observed by Ott et al (4) and Manzano et al (5), which could be explained by the higher amount of concentrate provided in the earlier studies.
Although the GE digestibility coefficient we observed was lower than that reported in other studies using citrus pulp (5)(6), the values did not differ depending on the levels of citrus pulp included (Table 4), indicating that the observaron valores de digestibilidad aparente similares.Las diferencias entre los resultados se pueden atribuir a las distintas calidades de forraje utilizadas en los estudios.Manzano et al (5) suministró alimentos una relación forraje:concentrado de 40:60 ad libitum y observó coeficientes de digestibilidad de nutrientes más altos que los que observamos nosotros.
Aunque el coeficiente de digestibilidad de EB que observamos fue menor a aquel reportado en otros estudios que utilizaron pulpa de cítricos (5)(6), no hubo diferencia en los valores de acuerdo a los niveles de pulpa de cítricos incluidos (Tabla 4), lo cual indica que la inclusión de este ingrediente no inclusion of this ingredient did not negatively alter the digestibility of GE.The consistent GE digestibility coefficients can also be explained by the consistent levels of NDF and ADF digestibility (2), which provided a constant substrate for degradation by microorganisms, which produces energy.
Adding fiber to the diet increases intestinal motility, which may impair the digestibility of nutrients in the small intestine.This could explain the low value of the GE digestibility observed in this study.
The presence of fiber as well as the quality and quantity of food intake, particle size and speed of transit of the digesta can affect both nutrient digestibility and feces quality in horses (21).
Studying nutrient digestibility in conjunction with the physicochemical characteristics of feces is important for evaluating digestive tract function, as it allows the diagnosis of gastrointestinal disorders.
We observed no effect (p>0.05) of diet on the color of the feces, which had a greenish color.Feces color is primarily influenced by the forage used (22), and the color observed in this study was due to the characteristics and the amount of the hay provided (75% Coast Cross hay on a DM basis).
The consistency of the feces observed in this study was normal, suggesting that adding citrus pulp to the diet does not promote production of high levels of lactic acid and consequently feces fluidization, which can be proven by maintaining the pH and buffering capacity (Table 5).
The fecal pH was not influenced (p>0.05) by treatment (Table 5).The mean pH value we observed (6.62) is considered appropriate for the activity of cellulolytic bacteria that ferment cell wall carbohydrates, thereby producing short chain fatty acids (SCFA) that are absorbed in the large intestine (23,24).
The buffering capacity conferred the feces, may be related to ion exchange at the SCFA absorption time, which occurs through the absorption of sodium ions (Na + ) at the apical membrane of the absorptive cells in exchange with hydrogen ions (H + ), this process can be directly linked to the exchange of chlorine (Cl) and bicarbonate (HCO 3 ) (21), process according to Van Soest et al (15), is essential for maintaining proper pH levels, as evidenced in this study.alteró de forma negativa la digestibilidad de EB.La consistencia de los coeficientes de digestibilidad de EB también se puede explicar por los niveles consistentes de digestibilidad de FDA y FDN, que proporcionaron un sustrato para la degradación por parte de microorganismos, lo cual produce energía.
En este estudio se observó que la consistencia de las heces fue normal, lo cual indica que agregar pulpa de cítricos a la dieta no propicia la producción de altos niveles de ácido láctico y por ello la fluidización de las heces, lo cual se puede demostrar ya que se mantiene el pH y la capacidad tampón (Tabla 5).

Moreira -Palatability and digestibility of citrus pulp for horses
The forage:concentrate ratio (75:25, based on DM) may have created favorable conditions in the large intestine for cellulolytic bacteria, data that corroborate with Hoffman (25) findings, which cites that diets with higher amounts of roughage, the main SCFA produced is the acetate.
This indicates that the main substrates for fermentation in large intestine were rapidly and slowly fermentable carbohydrate, with predominance production of weak acids.
The quantity and the quality of fiber in the diet may modify intestinal transit, leading to changes in fecal characteristics.The treatment tested in this present study provided a high amount of fiber as a substrate for the microorganisms.The fiber also has high BC due to the exchange of cations and metallic ions (16).Consistent with this concept, we observed no reduction in BC in this study (Table 5), this demonstrates that animals fed diets with increasing levels of citrus pulp -an ingredient that is considered a "super fiber"-maintained a fecal pH of above 6.0, which is optimal for fermentation by cellulolytic bacteria.
In conclusion horses can identify the presence of citrus pulp in concentrate and prefer concentrates without citrus pulp.
Adding up to 28% citrus pulp to horse diets does not negatively affect the digestibility of nutrients and the physicochemical characteristics of the feces; thus, it is an alternative energy ingredient in the formulation of horse diets.
Agregar un 28% de pulpa de cítricos a las dietas de los caballos no afecta de forma negativa la digestibilidad de los nutrientes ni las características fisicoquímicas de las heces, y por lo tanto es una alternativa de ingrediente energético en la formulación de dietas para caballos.

Table 1 .
Percentage composition of the experimental concentrates with increasing levels of citrus pulp added to the diet.

Table 2 .
Chemical composition of concentrates and experimental diets with increasing levels of citrus pulp added to the diet.

Table 3 .
First choice and intake ratio of concentrates with increasing levels of citrus pulp added to the diet.
dLowercase letters indicate significantly different pairs, as determined by Student's t test (p<0.001).

Table 4 .
Digestibility coefficients of the nutrients of diets with increasing levels of citrus pulp added to the diet.
1No significant difference in the mean, as determined by Tukey test (p>0.05).* CV -coefficient of variation.

Table 5 .
Buffering capacity and pH of feces from horses fed on diets with increasing levels of citrus pulp.