Metabolisable energy levels for stress broilers under thermal stress

The research was carried out with the objective of evaluating different levels of metabolisable energy (3050, 3100, 3150, 3200 and 3250 kcal.kg–1) in the diet of 22-to-42-day-old broilers kept in an environment of thermal stress. Three hundred male chicks weighing 814.7 g initial average weight were distributed in a completely randomised design with five treatments, six replications and ten birds per experimental unit. the performance variables of feed conversion, daily weight gain and daily feed intake in the two phases studied. There were no statistical differences in carcass yields: carcass, chest, thigh, drumstick, heart, wing and back. Differences were observed for the two variables gizzard yield and abdominal fat. A linear increase in the percentage of abdominal fat with an increase in energy levels in the diet was observed. The 3100 kcal.kg–1 level of metabolisable energy is the level indicated by this research because it results in lower abdominal fat content and lower cost per kilo of feed.


INTRODUCTION
The Brazilian agricultural sector has transformed poultry farming into an industrial enterprise, as well as a great source of animal protein production. According to the Brazilian Association of Animal Protein ABPA (2018), Brazil is the second largest chicken meat producer in the world, with production of 13.06 million metric tons in 2017. Of the total volume produced, 66.9% was for domestic consumption and 33.1% for export. Of the total domestic consumption, 34.32% was slaughtered in the state of Paraná, the state that slaughtered the most (BAPA, 2018). Broiler chickens are homeotherm animals; that is, they can maintain body temperature within a narrow range when subjected to variations in the thermal environment within a certain range. To do so, they may use behavioural adaptations such as wing spreading, spreading or nesting relative to other birds (SCHIASSI et al. 2015). According to Silva et al. (2012), the process of maintaining homeothermia is only efficient when the ambient temperature is within the limits of thermoneutrality, and birds do not adjust perfectly to temperature extremes, and may even have their lives threatened. The high growth rate of modern broiler chickens is largely the result of genetic improvement and production conditions such as nutrition and management (RAMOS et al., 2011). Thus, knowledge of the chemical and energetic composition of the ingredients used in poultry diets is necessary for the formulation of more efficient and economical rations, since food is the most expensive item in the production of broiler chickens (CAIRES et al., 2008). Oils and fats have emerged as excellent sources of energy and essential fatty acids in poultry feed, increasing feed energy intake to compensate for low nutrient intake due to reduced consumption in warm climates and improving palatability (NOBAKHT et al., 2011). Broiler breeders in the semi-arid regions point to difficulties in formulating feed. According to Medeiros et al. (2015) tables direct the fulfilment of the nutritional requirements for birds in temperature regions and mild environments, since these tables were developed based on data from experiments performed in these particular regions. Based on this, the nutritional requirements of the birds are modified in high-temperature situations. Thus, the objective of this research was to evaluate the performance of broiler chickens aged 22 to 42 days fed diets containing different levels of metabolisable energy in hightemperature environments.

MATERIALS AND METHODS
The experiment was conducted in the poultry sector of the and end of the experiment, the feed was given ad libitum, and the dietary leftovers were collected at the end of each phase, thus obtaining data for the calculation of daily weight gain (DWG), feed conversion (FC) and daily feed intake (DFI). The animals were slaughtered at the age of 42 days. After slaughter, the birds were weighed for carcass evaluation. Carcass, chest, thigh, drumstick, liver, heart, gizzard, wing, back and abdominal fat were evaluated. The parts were weighed on a precision scale and data collected for analysis.
The results were submitted to variance analysis and when there was significant difference, regression analysis was applied using the SISVAR software. The means of the treatments were compared by Tukey test at 5%.

RESULTS AND DISCUSSION
The results for DWG, DFI and FC of broilers fed with different levels of metabolisable energy in the growing phase and maintained in a thermal stress environment are given in Table 2. The DWG of the birds was not influenced (P > 0.05) by the dietary metabolisable energy levels in the growing phase. These results are dissimilar to those reported by Barbosa et al. (2008), who used metabolisable energy levels in diets for broilers kept in a high-temperature environment (maximum average temperature of 36.6 ºC), presenting the highest absolute weight gain of 61.72 g in the growth phase when a level of metabolisable energy (ME) of 3100 kcal.kg-1 was used.
In the present research, no significant differences (P > 0.05) were observed for FC in the growth phase. These results differ from those reported by Oliveira Neto (2000), who evaluated the effect of ambient temperature on the performance and carcass characteristics of broilers fed a controlled diet with two levels of metabolisable energy: increasing the level of ME in the diets from 3075 to 3300 kcal.kg-1 resulted in improvements in FC from 2.27 to 2.25 (g.g-1). When exposed to high ambient temperatures, birds decrease food consumption substantially (BOIAGO et al., 2013). Consequently, they increase their water intake, as reported by Laganá (2008), which results in a worse FC index. The results found in this research in the growth phase agree with those of Cordeiro et al. (2010), who found the best FC value of 1.68 g.g-1 in the third week of the birds' life. These same authors state that broilers under heat stress reduce feed intake and increase energy expenditure for heat dissipation, resulting in high FC and low weight. The DFI of the birds was not influenced (P > 0.05) by the ME levels of the diets in the growing phase. These results corroborate those of Sakomura et al. (2004), who evaluated the effect of dietary ME level on performance and energy metabolism of broiler chickens and observed that the ME levels of the diet did not affect the DFI: 171.43 g for birds that consumed 3200 kcal.kg-1. The stress generated by the variation in the thermal environment experienced by the animal directly influences productivity by altering the heat exchange with the environment, the amount of food consumed, the body weight gain and, consequently, the nutritional requirements (SOUZA & BATISTA, 2012). No significant differences (P > 0.05) were observed between treatments in the finishing phase for broilers fed with different energy levels, as indicated in Table 3. For DFI no statistical differences (P > 0.05) were observed between treatments. Lima et al. (2008), in a study with different energy levels of 2900, 3000 and 3100 kcal.kg-1, lysine and methionine + cystine, in the diet for male broilers from 1 to 40 days, observed that the lowest daily intake of 100.57 g was obtained with the lowest energy level (P < 0.05). For these authors, higher intake may be associated with higher oil levels in medium-and high-energy diets; this increases palatability, favouring increased feed intake. This behaviour was not confirmed by the results obtained in this study.
In the present research, no significant differences in DWG (P > 0.05) were observed for different ME levels of the diet in the finishing phase. Similar results were found by Mendonça et al. (2008), who observed no effect of feed energy levels on DWG, evaluated in their research metabolisable energy levels for slow growing male broilers raised from 22 to 49 days of age, found values for a greater DWG in the treatment that was offered 3150 Kcal.kg-1 of MS obtained 36.21 g, although the lowest value found in absolute weight was 34.39 g for the 2700 Kcal treatment. kg-1. The nonsignificance in the results of this research can be justified by the fact that the ambient temperature is determinant in the performance of broilers, a physical factor that has a great influence on the Rev. Bras. Saúde Prod. Anim., Salvador, v.21, 01 -13, e210422020, 2020 ISSN 1519 9940 http://dx.doi.org/10.1590/S1519-9940210422020 Animal Production and Environment Received on: 03/09/2019 Accepted on: 19/11/2019 6 amount of food consumed by the birds, directly compromising the weight gain. According to Nazareno et al. (2009), the thermoneutrality zone is related to an ideal thermal environment, in which birds experience suitable conditions to express their best productive characteristics. No significant differences (P > 0.05) were observed in FC in the termination phase with the ME levels of the diet. These results confirm those obtained by Roll et al. (2011), who also found no effect of energy levels on FC, obtaining a better value of 2.44 g.g-1 for this variable when a diet of ME level 3050 kcal.kg-1 was offered between 43 and 48 days of age. According to Olanrewaju et al. (2010), processes such as panting and wing opening to dissipate heat require extra energy expenditure. Therefore, at high temperatures there is a reduction in food utilisation efficiency, resulting in a generally higher FC rate. In this study, the average temperature was 30.3 ºC during the growth phase and 26.1 ºC during the termination phase, characterised as outside the thermal comfort zone in both phases. According to Tinôco (1998), an environment considered comfortable is 16-23 °C. Thus, these average temperature values during the research may have needed greater heat dissipation by the birds to remain in thermal equilibrium. For Beterchini et al. (1991), better results are obtained in broilers in the final phase in an environment with an average temperature of 22 ºC, with no significant differences seen between 17.1 and 27.9 ºC. The relative humidity of the air during the growth and termination phase was 49.0% and 61.2%, respectively. According to Naas (1995), the appropriate relative humidity range for broiler chickens in their fifth week should remain in the range of 50-70%. The author reported that very high temperatures associated with high relative humidity values cause a reduction in productive performance and economic losses, especially in the final stages when chickens are nearing the age for commercialisation. For carcass yield, the variables abdominal fat and gizzard yield were significantly different (P < 0.05) between treatments, as shown in Table 4. In this research no significant differences (P > 0.05) were observed in carcass yield for the different ME levels of the diet; the highest relative weight of 80.31% was seen in the treatment with 3250 kcal.kg-1 ME. Similar results were observed by Meza et al. (2015), who used ME and digestible lysine levels for broiler composition and carcass yield and did not notice significant differences on carcass yield, obtaining the highest relative weight of 73.20% when offering a diet of 3360 kcal.kg-1 ME. According to Baracho et al. (2013), it is known that Rev. Bras. Saúde Prod. Anim., Salvador, v.21, 01 -13, e210422020, 2020 ISSN 1519 9940 http://dx.doi.org/10.1590/S1519-9940210422020 Animal Production and Environment Received on: 03/09/2019 Accepted on: 19/11/2019 8 animals reach their optimum productivity when kept in a thermoneutral environment; that is, when food energy is not diverted to compensate for deviations from the thermoneutral range by eliminating or maintaining heat. Significant differences (P < 0.05) were found for the variable abdominal fat, and it is possible to observe a simple linear effect in the regression analysis. The best index for this variable is for the 3050 kcal.kg-1 ME diet and the worst for 3250 kcal.kg-1 according to the equation y = 0.006032148x -15.551953554; obtaining 2.84% abdominal fat for treatment with 3050 kcal.kg-1; see Fig.  1. Birds showed increased abdominal fat deposition when higher energy levels were provided in the diet, with the largest deposition of 4.05% at 3250 kcal.kg-1 ME. This disagrees with the findings of Duarte et al. (2012), who evaluated the performance and duodenal morphometry of broiler chickens submitted to different energy levels and feeding programmes from 42 to 57 days old, in which the birds' abdominal fat percentage was not affected (P > 0.05 ) by the energy levels of the ration, levels of 3200 and 3600 kcal.kg-1 obtaining values of 2.00% and 2.14%, respectively. According to Oliveira Neto et al. (2000) this may be associated with the reduced size of metabolically active organs and protein deposition in the breast observed in birds maintained at 32 ºC which provided greater amount of energy for deposition as fat in the heat environment. In addition, if the source of fat used to formulate the diet is rich in polyunsaturated fatty acids, such as soybean oil, this contributes to a greater accumulation of abdominal fat (LOPES et al. 2015).
For the variable gizzard yield, the different energy levels exerted an influence (P < 0.05). In this research the highest gizzard yield was in the treatment with 3100 kcal.kg-1 ME, obtaining 1.54% relative weight and the lowest gizzard yield with 3050 kcal.kg-1 ME, with a value of 0.95 relative weight. A quadratic effect was observed in the applied regression analysis; according to the equation y = -0.000032942x2 + 0.208339879x -327.939640498 it can be observed that the ideal level is 3162.2 kcal.kg-1 ME, yielding 1.46% -see Fig. 2. In contrast, these results do not corroborate those found by Barbosa et al. (2008), who reported that ME levels did not influence the gizzard performance of birds, evaluating effects of ME levels of 2800, 2900, 3000, 3100 and 3200 kcal.kg-1 in the final phase 22 -49 days), obtaining the highest relative gizzard weight of 2.95% with 3200 kcal.kg-1 treatment at 49 days of age, and kept in a hightemperature environment. This value, which is 2.95% higher than the value found in this research, can be justified by the 7 days longer that this experiment was conducted, which, in the case of broilers, may be the number of days required for several physiological changes to occur in birds. The values found in this research for the variable gizzard yield (Fig. 2), corroborate those found by Oliveira Neto et al. (2000), who obtained a relative weight of 1.59% when using a diet of ME level 3075 kcal.kg-1, evaluating the effect of room temperature on the performance and carcass characteristics of two-diet controlled broiler chickens. These authors reported that the observed reduction in the weight of the organs of birds exposed to high ambient temperatures is a physiological adjustment to reduce body heat production. Taking into account the differential of feed cost reduction, as indicated in Table  5, it is noted that the feed with the highest cost in each phase (growth and termination) was found for treatment with 3250 kcal.kg-1 of , obtaining values of R$ kg-1 1.96 and 1.92, respectively. However, the lowest costs y = -0,000032942x 2 + 0,208339879x -327,939640498 R² = 0,632053277 Gizzard (%) Energy level (Kcal Kg -1 ) Rev. Bras. Saúde Prod. Anim., Salvador, v.21, 01 -13, e210422020, 2020 ISSN 1519 9940 http://dx.doi.org/10.1590/S1519-9940210422020 Animal Production and Environment Received on: 03/09/2019 Accepted on: 19/11/2019 10 (R$ Kg-1 1.86 and 1.82, respectively) were offered by the rations with a level of 3050 kcal.kg-1 ME.