ESTIMATION OF THE REQUIREMENT OF METABOLIZABLE ENERGY AND CRUDE PROTEIN FOR GROWING IN RHODE ISLAND RED CHICKENS , UNDER TROPICAL CONDITIONS OF SOUTHEASTERN MEXICO

The objective of the current study was to estimate the requirement of metabolizable energy (ME) and crude protein (CP) for growing Rhode Island Red male chickens from 17 to 119 days-old, under tropical conditions of Yucatan, Mexico. The experiment consisted in three feeding phases (17–56, 57–91, 92–119 days old), using a factorial 2x3 design for each one. Two constant ME (2900 and 3100 kcal/kg DM), and three CP concentrations (% DM) that varied with age: 22, 20 and 18% (17–56 days-old); 20, 18 and 16% (57–91 days-old); and 18, 16 and 14% (92–119 days-old) were used. Two batches of 72 Rhode Island Red male chickens each one was used; the first one for the first feeding phase and the second batch for the second and third feeding phases. Response surface procedures to obtain the best ME and CP concentrations were carried out. The body weight gain (BWG) and feed intake (FI) were greater in the first feeding phase with 2900 kcal ME/kg DM in the diet; however, in the second phase there was not significant effect of energy concentration. From 17–91 days-old, the FI increased with 2900 kcal ME/kg DM in the diet. Chickens had a higher BWG from 92–119 days-old when 16% CP was included in the diet. Protein efficiency was better from 17– 119 days-old when the lower CP concentration was included in the diet. In conclusion, the greatest BWG were obtained, when 2900 kcal ME/kg DM, for the three feeding phases, was used. The best approximations of CP requirement in the diet for the first, second and third feeding phases were 21.3%, 19.2% and 17.1%, respectively.


INTRODUCTION
Slow growing poultry are commonly used in backyard poultry farming as they allow resource-limited people to produce their food (Centeno et al., 2007).In addition, this poultry type is used in free range and semi-intensive grazing systems because they have some characteristics such as survive in adverse circumstances, low appetite and adaptable to lownutrient density diets that make them easily adaptable to the climatic conditions of the tropics (Rizzi et al., 2013).
There are several reports showing that slow-growing chickens are less efficient in meat production (Almasi et al., 2015;Rizzi et al., 2013;Wang et al., 2013) but their meat may be healthier compared to commercial poultry meat because it contains less fat and more polyunsaturated fatty acids than medium-and rapidgrowing broilers (Martínez-Pérez et al., 2017).
Price of crude protein (CP) and metabolizable energy (ME) sources for animal diets have been increasing, which affect the cost of meat animal production (Ren et al., 2013;Rosegrant et al., 2013).Therefore, determining nutrient requirements and adequate feed formulation is a critical issue, because there is a direct relationship between the animal performance and the nutrient density in the feed (Horsted et al., 2010).Furthermore, slow growing poultry does not seem to require a high concentration of CP and ME in the diet for growing and development (Sundrum et al., 2006).However, despite the National Research Council (NRC, 1994) proposes requirements of ME and CP for broilers, there are no consistent reports of requirements for slow-growing breeds, such as Rhode Island Red type of chickens (Wolde et al., 2011b).Thus, it is important to know the energy and protein requirements of slow growing breeds birds with the purpose of implementing feeding strategies to take full advantage of the amount of nutrients offered in the diet.The objective of the current study was to estimate the requirement of ME and CP for growing Rhode Island Red male chickens from 17 to 119 day-old, under tropical conditions of the southeast of Mexico.

Study site
The experimental work was carried out from July to November 2016, at the animal nutrition area of the Faculty of Veterinary and Animal Science, University of Yucatan, Mexico (FMVZ-UADY).The climate in the region is subhumid with an average annual rainfall of 1100 mm; rainfall mainly occurring during summer (June to October).The average annual temperature is 26°C, with average maximum of 36 °C in May, and average minimum temperature 16 °C in January.

Animal management
Two batches of one-day-old Rhode Island Red male chickens were bought.The first batch was used for the first feeding phase (17-56 days-old) and the second batch for the second and third feeding phases (57-91 and 92-119 days-old, respectively).Both batches arrived in July 2016.At the arrival of the first batch, the chickens were weighted and assigned into six groups with similar body weight means.Chickens were kept in floor cages from 0-16 days old and each group was fed one of the six experimental diets used in the first feeding phase.When the chickens were 17days-old, they were individually allocated in elevated wired cages (40 x 40 cm), using a completely randomised design.
At the arrival of the second batch, the birds were randomly assigned to floor cages.The chickens remained in floor cages during 7 weeks.At week 8, they were weighted, individually allocated in elevated wired cages and assigned to six groups with the same means to start the test at week 9 of age (57 days-old).The diet given to the chicken in this period contained 20% CP and 3100 kcal ME/kg DM.In the third phase, the same chickens of phase two were used.The chickens were weighted, and body weight gain was recorded weekly.The floor cages used in both batches were equipped with a push-fit drinker and tray feeders during the first week of age.Automatic bell drinkers and hanging feeders were used later on.

Characteristics of diets
Three phases were assigned in the test with a different feeding program for each one.The feeding program consisted of two levels of ME (2900 and 3100 kcal/kg DM), which were constant along the test and three amounts of CP that decreased by 2% from one phase to another.In the first phase, 18, 20 and 22% CP in diets were used.In the second and third phases, 16, 18 and 20% and 14, 16 and 18% of CP in the diets were used, respectively.Lysine and methionine percentage were balanced according to broiler requirements (NRC, (1994), maintaining an approximated ratio of 2:1 (100:46).Soybean meal and maize were used as the main ingredients for diet formulation.The feed was prepared on site (experimental farm) and samples were taken for chemical analysis at the animal nutrition laboratory, FMVZ-UADY (Table 1).Chickens were offered feed ad libitum every day and the leftovers were weighed weekly to calculate the cumulative intake.The ME intake, CP intake, protein efficiency ratio (PER, gain per gram of protein intake) and energy efficiency ratio (EER, gain per 1 kcal ME intake) were also determined.At the end of the third phase, animals were sacrificed at the slaughterhouse of the FMVZ-UADY, and carcass yield was expressed as the percentage of final live weight when abdominal fat and nonedible components were removed (head, feet, liver, gizzard, heart, blood, feathers and gastrointestinal tract).

Statistical analyses
The experiment consisted in a completely randomized design with a factorial arrangement of treatments (12 replicates per treatment).The body weight gain (BWG), feed intake (FI), feed conversion (FC), crude protein intake (CPi), metabolisable energy intake (MEi), protein efficiency ratio (PER), energy efficiency ratio (EER), carcass performance and abdominal fat, were analysed through analyses of variance with the statistics package SAS version 9.4.
The mathematical model was: Where: Yijk is the ijk-th observation in the i-th level of factor A (ME) and j-th level of factor B (CP); µ is the general mean; αi is the effect of i-th level of factor A; βj is the j-th level of factor B; αβij is the interaction of i-th level factor A with the j-th level of factor B; and eijk is the random error.
Since BWG was the unique variable for which requirements were obtained in each feeding phase, it was analysed to determine the best response level, using the RSM tool of the statistics package Minitab® 17.The mathematical model was: Where: Y is the response of interest (BWG); β0 is the intercept and βi linear effect of factor A on BWG; βj is the linear effect of factor B on BWG; βii is the quadratic effect of factor A on BWG, βjj is the quadratic effect of factor B on BWG; βij is the interaction of factor A with levels of factor B on BWG; ai is the effect of level factor A; bj is the level of factor B; abij is the interaction of level factor A with level of factor B; and eijk is the random error.

RESULTS
Higher BWG, FI and FCR, were obtained with a diet containing 2900 kcal ME/kg DM compared to the 3100 kcal ME/kg DM diet from 17 to 56 days-old in chickens.The birds that consumed 2900 kcal ME had a higher CP intake and ME intake but less PER (Table 2).From the period 57-91 days-old, chickens that consumed 2900 kcal ME had higher FI, CP intake and EER but same FCR, compared with those consuming 3100 kcal ME in the diet (Table 3).From the time between 92 to 119-days-old, higher EER with 2900 kcal ME was observed (Table 4).
From 92-119 days of age, CP had a significant effect on BWG, where the highest response, was observed with 16% in the diet.Highest CP intake was observed with 16 and 18% CP in the diet, while 14% CP resulted in better PER.From the time the birds were 57-91 days-old, 18 and 20% CP had the lowest FCR and highest EER, whereas with 18 and 16% CP, birds had the best PER.CP intake was significantly higher with higher CP levels in the diet (22, 20 and 18%) for each feeding phase (17-56, 57-91 y 92-119 days of age, respectively).Nevertheless, the PER was lower with these CP levels in the diets.A significant interaction of CP intake and ME intake as factors was observed for FI, from 57-91 days old (Figure 1, 2 and 3).There was no effect of the main factors or interaction between them, neither for carcass yield nor for abdominal fat.
Table 5 shows the response surface for BWG in the three feeding phases.The BWG response surface for the second feeding phase showed that CP had neither linear nor quadratic effect and ME concentration had no effect on BWG.
Response surface analysis indicates that the optimum level for both CP and ME were 20.7% and 2900 kcal/kg, respectively, to reach 811 g BW.The BWG response was explained by the following model: BWG: 1116 -0.1577 (ME) + 7.37 (CP) In the response surface for the third feeding phase it was observed that the CP had a quadratic effect while ME had no effect.The response surface analysis indicates that the optimum amount for CP in the diet is 17.1% and 2900 kcal/kg ME to reach 852 g BW.The model explaining the BWG response was: The results of the current study are within the range of previous reports, considering that in the first feeding phase (17-56 days old) the RIR chickens had highest BWG consuming 2900 kcal ME/kg DM.This is attributed to the higher feed intake and consequently higher CP and ME intakes in comparison to those chickens whose diet contained 3100 kcal ME/kg DM, which, in an attempted to regulate their ME intake, consumed less ME and therefore, had lower BWG.Zhao et al. (2009) and Wang et al. (2013) also described a positive relationship between high feed intake and high BWG in chickens.
In the last two feeding phases (57-91 and 92-119 days old), there was no effect of dietary ME concentration on the BWG, in agreement with Raju et al. (2004), Haunshi et al. (2012) and Abudabos (2014).However, in the second feeding phase (57-91 days old), the birds significant decreased feed intake when they consumed 3100 kcal ME/kg DM, causing both lower CP intake and EER, compared to those that consumed 2900 kcal ME/kg DM in the diet.
In the period of 92-119 days of age, EER was significantly better with 2900 kcal ME/kg DM in the diet, although the BWG and feed intake were similar.This could be due to the high ME: CP ratio in the diet with 3100 kcal ME/kg DM.
The PER result was lower when the diet contained 2900 kcal ME/kg DM compared to 3100 kcal, probably due to the low caloric content in the first diet, inducing the chickens to consume more feed; therefore, BWG: CP intake ratio was higher compared to those chickens consuming diets with 3100 kcal ME/kg DM.The differences in the feed intake from the current study indicate that it depends on the concentration of ME in the feed, because the chickens eat to meet their energy needs (Haunshi et al., 2012;Iqbal et al., 2014;Rama Rao et al., 2014).Carmona et al. (2016) mentioned that abdominal fat increases as the age advances; in addition, Raju et al. (2004) andAkbari et al. (2016) reported that this increment is directly proportional to the ME in the diet.
The carcass yield weight also increments when both age advances and ME increases in the diet.However, from the time chickens were 92-119 days-old, the phenomenon, neither for abdominal fat accumulation nor for carcass yield was observed in the current study; this may be due to the fact that there were no differences in the ME intake.

Effect of crude protein
In slow-growing chickens, Mosca et al. (2016) recommended 16% CP in the diet for birds from 0 to 180 days-old; while Li et al. (2013) mentioned that 17% CP should be used from 0 to 60 days-old, and Haunshi et al. (2012) reported 18% CP from 0 to 56days-old.These recommendations indicate that the requirement of CP in slow-growing chickens could be between 16 and 18% when given a single concentration during the growth period.
However, the current study used three schemes where the CP in the diet ranged between 14 and 22%.Similarly, Wolde et al. (2011) used five diets (14,16,18,20 and 22% CP) for 1-91 days-old RIR chickens and recommended 16% CP in the diet.This result differs with the current study, in which the chicken growth was divided in phases because the requirement of nutrients is higher in the early age; likewise, it is related to the rate of protein accretion in the body that is higher in the first two weeks of life (Kang et al., 1985;Wen et al., 2016).
The response surface analysis (RSM) yielded intermediate levels (21.3, 19.2 and 17.1% CP per phase) as ideal to improve the BWG response.In contrast, when using the ANOVA tool, the recommendations varied (18, 16 and 16% CP).The use of RSM was considered since it adjusted the behaviour of the dependent variable per the independent variables, when performing a multiple regression analysis (Khuri et al., 2006;Mehri et al., 2012).
Using the RMS (from 92-119 days-old), it was found that the CP level in the diet had a quadratic effect on the BWG.A difference of 27 g CP intake between both diets caused a significant decrease of 73 g BWG, even though, the CP intake had no significant effect on BWG when diets had 16 and 18%.This difference could be due to an excessive CP intake, since there is a direct relationship between CP intake and uric acid level in blood, derived from the amino acid catabolism of the diet (Hernández et al., 2012).According to Buyse et al. (1992) there is greater nitrogen excretion or fat deposition through lipogenesis which leads to a higher energy expenditure affecting the growth performance of chickens.
In the current study, the CP in the diet had no effect on the feed intake in agreement with From 57-91 days-old, the EER and PER were higher with 18% CP in the diet.It means that with 100 kcal ingested, there will be 13.6 g BWG and for every 100 g CP intake, there will be 227 g BWG.However, with a significant decrease of 16% in the EER, 20% of the PER decreased.This is due to the CP intake that directly affects these measurements (Cheng et al., 1997;Haunshi et al. 2012).
In the current study, no effect of dietary CP on abdominal fat was observed, in agreement with Awad et al. (2014) and Wolde et al. (2011).Although in the third phase (92-119 days old) there was a quadratic effect of CP on GDP, this suggests that the N from amino acids could be eliminated by the excreta as reported previously by Hernández et al., (2012).However, the number of observations in our study was not enough to show significance in the detected trend (13 g/kg abdominal fat difference).

Interaction of factors
The interaction of the main factors (ME and CP) was significant on FI, CP intake and ME intake during the second feeding phase (57-91 days old).The best response was obtained with 20% CP and 2900 kcal in the diet.On the other hand, the chickens that consumed the diet with 16% CP and 3100 kcal, had a high feed intake.However, the lower CP intake obtained was not enough to cover the requirement for 57-91 days and the ME intake was very high, which could contribute to the greater deposition of body fat (McLeod et al., 1991;Zeng et al., 2015), taking into account that fat synthesis requires higher ME than protein synthesis.
With the diets of 18% CP, the concentration of ME had no influence on the CP intake, consuming similar amounts of CP as the 20% diet with 3100 ME.This last diet had high nutrient concentration, low feed intake, high CP intake and low ME intake.The imbalance in nutrient intake could cause adverse effects on the growth rate (Kamran et al., 2008) an effect observed in the response surface analysis trend.

CONCLUSION
The best dietary level of ME requirement for the three feeding phases was 2900 kcal/kg to obtain greater BWG.On the other hand, 3100 kcal/kg ME decrease FI and FCR, but BWG was negatively affected.The carcass yield and abdominal fat content were not affected with the CP and ME levels tested.An approximation of CP requirement in the diet for the first, second and third feeding phase would be 21.3%, 19.2% and 17.1%, respectively.

BWG
Some recommendations regarding dietary ME concentrations in slow-growing chickens and their effect on BWG are described in the literature.Nahashon et al. (2006) reported 3100 kcal ME/kg DM;Raju et al. (2004) andLi et al. (2013) mentioned 2900 kcal ME/kg DM andHaunshi et al. (2012) found that 2800 kcal ME/kg DM was sufficient to maximise the BWG.These reports indicate that the range of ME for slow-growing chickens could be between 2800 and 3100 kcal ME/kg DM.The differences reported may be attributed to the particular growth rate of the genotypes used in each study.

Table 1 .
Analysed and calculated values of the ingredients of experimental diets.

Table 5 .
Orthogonal analysis of metabolizable energy (ME) and crude protein (CP) effect on body weight gain.
Figure 1.Interaction of crude protein and metabolizable energy for feed intake since 57-91 days old.Figure 2. Interaction of crude protein and metabolizable energy for crude protein intake since 57-91 days old.Figure 3. Interaction of crude protein and metabolizable energy for metabolizable energy intake since 57-91 days old.For the BWG response surface of the first feeding phase CP had neither linear nor quadratic significant effect; however, ME had a linear effect.The response surface analysis indicates that the optimum level for CP in the diet was 23.0% and 2900 kcal/kg ME to reach 862 g BWG.The model explaining the BWG response was: BWG= 1843 -0.3708 (ME) + 4.11(CP) Cheng et al. (1997))11)andZeweil et al. (2011), feeding laying hens, andWolde et al. (2011)feeding RIR chickens.They found high CP intake when the diet had high CP concentration.From 57-91 days, FCR significantly worsened with the 16% CP level in the diet, associated with a low CP intake in the low dietary levels and vice versa.Cheng et al. (1997)andBregendahl et al.