Nutrition is one of the main factors for the success of any segment of animal production, being fundamental to obtain high production indices. In this context, mineral supplementation, which has been used in broiler farming for promoting improvements both in performance and the main carcass characteristics ( Zia et al., 2017 ) stands out.

Among the minerals used in the broiler diet, we highlight both chromium and selenium, which are essential for animal development and perform various metabolic functions in the organism. Chromium acts as an active component of glucose tolerance factor (GTF), promoting greater insulin binding to receptors, thus increasing the sensitivity of target cells and, consequently, enhancing insulin action ( Carvalheira et al., 2002 ). Selenium promotes increase in the circulating levels of triiodothyronine (T3) due to its action on the activation of the enzymes iodothyronine deiodases ( Ahmad et al., 2012 ). Thus, supplementation with both minerals in the diet may lead to increased protein synthesis in muscle tissue, which may be responsible for performance improvements and increments in the main carcass characteristics ( Silva et al., 2019 ; Ognik et al., 2020 ). However, for the inclusion of these minerals in the broiler diet, it is necessary to assess the economic viability of the supplementation to verify whether the possible gains in performance and/or in carcass characteristics are economically advantageous.

The real effects of chromium and selenium supplementation in biocomplexed forms on broiler performance and carcass parameters are not yet fully understood. In addition, there are no recent studies in the literature that evaluate the economic viability of supplementation with these minerals and their influence on the growth pattern of broilers.

In view of the exposed, we aimed to check the influence of the inclusion of biocomplexed chromium and selenium in the diet on the performance and carcass characteristics of Label Rouge broiler chickens of both sexes raised in an alternative system, as well as to evaluate the economic efficiency of the mineral supplementation and to estimate the costs with the feeding of the birds according to the experimental diets.

The study on animals was conducted according to the institutional Ethics Committee on Animal Use (Protocol Number 29A/2016). The experiment was conducted in Bambuí, Minas Gerais. The municipality is located in the midwestern region of the Minas Gerais state, at an altitude of 706 m, with geographic coordinates 20°00'23’ South latitude and 45°58'37’ West longitude of Greenwich.

The experimental design was completely randomized arranged in a factorial scheme (3×2), namely, three experimental diets (control - without the addition of biocomplexed minerals; 0.50 ppm of biocomplexed selenium; and 0.40 ppm of biocomplexed chromium) and two sexes, amounting to six treatments. Each treatment consisted of three plots, each one represented by 21 birds, totaling 63 birds per treatment. Therefore, 378 birds were used for evaluations of performance and carcass. The experimental diets were offered throughout the production period, that is, from 1 to 90 days of age. In the diet with chromium, 0.34 g of chromium picolinate was added to 100 kg of feed (concentration of biocomplexed chromium of 12% in the product and 98.64% purity), and for the treatment with selenium, 5.10 g of selenium glycinate to 100 kg of feed (concentrations of 0.98% biocomplexed selenium in the product). The diets between the different treatments were isonutritive. In all the experimental diets, a vitamin premix with inorganic selenium was used, thus providing 0.37 ppm of inorganic selenium in the diet. In addition to the inorganic selenium, there was extra supplementation of biocomplexed selenium and chromium in two of the experimental diets.

The mathematical model for statistical analysis was as follows:

in which Y_ikl = effect of mineral supplementation i, at sex level k, on repetition l; µ = constant associated with all data; M_i = effect of mineral supplementation (i = 1, 2, 3); S_k = sex effect (k = 1, 2); MS_ik = effect of interaction mineral supplementation i and sex k; and e_ikl = error associated with all observations, which, by hypothesis, has a normal distribution with zero mean and sigma squared variance.

The diets were formulated based on corn and soybean meal according to the nutrient allowances of the Manual de Manejo Linha Colonial ( Globoaves, 2015 ), specific to the Label Rouge strain, to meet the requirements of the birds in each production phase ( Table 1 ). The feed compositions were taken from the Brazilian Tables for Poultry and Swine ( Rostagno et al., 2011 ). The calculated values of the experimental diets of each phase are presented in Table 2 .

The birds were raised in the period of 1 to 29 days (starter phase), in a conventional experimental shed, without access to grazing area, receiving heating by means of 250 W infrared lamps until 14 days old and artificial lighting 24 h a day; feed and water were supplied ad libitum . In the periods of 30 to 49 days (grower phase I), 50 to 77 days (grower phase II), and 78 to 90 days (finisher phase), the birds were reared in an experimental area of free-range poultry production, with access to the grazing area with Tifton 85 ( Cynodon spp ), both feed and water being supplied ad libitum . In each experimental unit of 77.35 m² of area, 21 birds of the same sex were housed, obtaining a density around one bird for each 3.68 m² of free area, discounting the coverage area of the mobile hut and thus meeting the Circular Letter DOI/DIPO number 007/99 of 19/05/1999 ( Brasil, 1999 ).

Regarding performance studies, the average weight, weight gain, feed intake, and feed conversion were evaluated in the periods 1 to 7, 1 to 21, 1 to 35, 1 to 49, 1 to 63, and 1 to 90 days, according to the methodology of Del Castilho et al. (2013) . Mortality was monitored daily for correction of feed intake and conversion, as reported by Sakomura and Rostagno (2007) . To assess the growth potential of the birds fed the experimental diets and of sex, the estimative of the parameters of the Gompertz equation, according to the methodology of Sakomura and Rostagno (2007) , was carried out from the data of average weight.

At the end of the experiment, five birds from each plot were selected for slaughter, with live weight close to the plot mean, varying 5% upwards or downwards. They were fasted for an 8-h period and slaughtered at 90 days of age. Slaughter was carried out in a slaughterhouse, and the birds were hung in the overhead rail, followed by sensitization by electronarcosis and bleeding. Afterwards, carcasses were scaldded, plucked, eviscerated, and cleaned. They were individually wrapped in duly identified plastic bags, following the distribution of the treatments and plots and refrigerated at 5 °C in a cold room for a period of 24 h for carcass analysis. During evisceration, edible viscera (heart, liver, and gizzard) and abdominal fat (retroperitoneal fat) were collected.

After the cooling period, the carcasses were weighed, considering the carcass with neck, head, and feet and free of abdominal fat and edible viscera. Next, the division of the cuts was carried out, separating the breast, thigh, drumstick, wing, back, neck plus head, and feet. For carcass analysis, live weight at slaughter, weight and carcass yield, the weights and yields of the above-cited cuts, the weights and yields of edible viscera (liver, heart, and gizzard), and abdominal fat were evaluated according to the methodology of Faria et al. (2011) .

To survey the economic viability of the mineral supplementation, the average cost (ACD) estimative, represented in Reais per kilogram of feed, of the experimental diets, by means of the mean of the costs of the diets in each production phase (starter, grower I, grower II, and finisher) was carried out. For this, a survey of the prices of the dietary ingredients was conducted during the experimental period, which was in March, April, and May of 2017, practiced in the state of Minas Gerais. Afterwards, the economic efficiency index (EEI) was calculated according to the methodology of Fialho et al. (1992) . The average feeding cost (AFC) was estimated during the whole production period (1 to 90 days), by means of the product between the average intake of the birds (kg/bird) and the ACD, the values being expressed in Reais per bird; and the cost per kilogram (kg) of weight gain (CWG) by means of the product between the ACD and the feed conversion, the values being expressed in reais per kg of weight gain.

The data were analyzed with the support of the SISVAR^® statistical program ( Ferreira, 2000 ). The variables with responses of significant effect in the analysis of variance for treatments and/or interactions were subjected to the Tukey average test (5% significance).

The birds fed the male sex-control diet presented greater estimated value of live weight at maturity (Pm). The birds of the chromium treatment and of the female sex presented higher estimated value of the maturity rate (b) and younger age at which the maximum growth rate (T*) is maximum ( Table 3 ).

The statistical analysis revealed no interaction between diet and sex (P>0.05) and isolated dietary effect (P>0.05) for all variables, regardless of the period ( Table 4 ). There was only an effect of sex (P<0.05) for the variables of average weight and weight gain, regadless of the periods, except for 1 to 7 days, and males presented the highest means. Regarding feed intake, males presented higher means (P<0.01) in the periods from 1 to 35, 1 to 49, 1 to 63, and 1 to 90 days. In addition, males presented better values (P<0.05) of feed conversion in the periods from 1 to 63 and 1 to 90 days.

There was no interaction between diet and sex (P>0.05) and any isolated effect of the diet (P>0.05) on the carcass variables, but there was sex influence (P<0.05) for all the variables, except for drumstick yield ( Table 5 ). Males presented higher values (P<0.01) of live weight, carcass weight and yield, breast weight, thigh weight and yield, drumstick weight, wing weight and yield, back weight, weight and yield of head plus neck, and feet weight and yield; however, females presented higher yield values of breast (P<0.05) and back (P<0.01) ( Table 5 ).

The statistical analysis revealed no interaction between diet and sex (P>0.05) for the variables of edible viscera and abdominal fat, but there was an isolated effect of the diet (P<0.05) for liver weight and yield ( Table 5 ). For both variables, birds fed the control diet showed a greater mean (P<0.05) relative to the selenium-fed birds; however, they were similar to those fed the chromium treatment ( Table 5 ). Regarding sex, there was an effect (P<0.05) for all the variables except gizzard yield. Males presented higher means of edible viscera weight (P<0.01), heart weight and yield (P<0.05), liver weight (P<0.05), and gizzard weight (P<0.01); females presented higher mean (P<0.01) of edible viscera yield, liver yield and weight, and abdominal fat yield ( Table 5 ).

The control treatment presented the best economic efficiency index (EEI), followed by the chromium treatment and, finally, the selenium diet ( Table 6 ). There was no interaction between diet and sex (P>0.05) and isolated effect of the diet (P>0.05) for any of the variables. Regarding sex, males presented higher value (P<0.01) of average feed cost (AFC) and lower value (P<0.05) of cost per kilogram of weight gain (CWG), besides the best EEI ( Table 6 ).

The value of weight at maturity (P_m) of the Gompertz equation is an estimative that indicates how much of live weight the birds in a given treatment will have when they reach maturity ( Sakomura and Rostagno, 2007 ). In this case, birds fed the control diet and males presented higher P_m value, which indicates that these birds may present higher values of live weight during maturity, when compared with birds of other treatments. Regarding sex, males tend to have higher live weight during maturity due mainly to the higher production of hormones with anabolic effects ( Santos et al., 2005 ).

The higher value of maturity rate (b) observed for birds fed the diet rich in chromium and females indicates that these birds have greater daily weight gain and, consequently, greater growth potential. In addition, they were younger for the maximum growth rate (T*), showing higher earliness for these birds. With advancing age, the growth rate decelerates, and the moment of change in the acceleration pattern of this curve characterizes the inflection point, corresponding to the value of T* ( Kessler et al., 2000 ). Females have higher earliness ( Santos et al., 2005 ), which makes them reach the maximum growth rate at a younger age. Coupled with this, chromium treatment for females increased maturity rate and reduced age for maximum growth rate, which can be attributed to the effect of this mineral as an insulin hormone enhancer ( Huang et al., 2016 ). The values of the Gompertz equation parameters of all treatments presented high reliability, with R² values above 0.98.

Supplementation with chromium and selenium minerals in broiler diets may cause physiological changes in poultry, which may generate performance response. Chromium has a reduction effect on the secretion of catecholamine ( Samanta et al., 2008 ) and corticosterone ( Sahin et al., 2003 ), which, when at high levels, may cause a reduction in feed intake ( Huang et al., 2016 ). Thus, chromium can increase feed intake and, consequently, weight gain. In addition, due to the fact that chromium acts as a potentiator of insulin action, it may lead to an increase in the utilization rate of glucose, which results in an improvement in feed efficiency, which may lead to an increase in weight gain and, consequently, an improvement in feed conversion ( Sahin et al., 2003 ). Selenium supplementation, in turn, may increase circulating levels of T3, which stimulates protein synthesis in muscle tissue and may be responsible for increased weight gain and improved conversion food index ( Ahmad et al., 2012 ).

However, in the present study, there were no changes in bird performance with supplementation with both minerals. This can be attributed to the fact that in the vitamin premix, used in the composition of the experimental diets, there was inorganic selenium. Therefore, this element was present in all the experimental diets, which may have interfered in the response of biocomplexed selenium supplementation, preventing changes in the performance of birds. Regarding chromium, this element is present in the main dietary ingredients such as corn and soybean, which would probably be enough to meet the nutritional requirements of the birds. In addition, since the birds were raised in a semi- intensive system with access to grazing areas, chromium and selenium could be present in the soil and, consequently, in plants. This may have increased the birds' intake of these minerals, preventing the response of supplementation with these minerals in biocomplexed form on performance. A great part of the works in the literature has also not observed the influence of the supplementation with chromium ( Perai et al., 2014 ; Zheng et al., 2016 ; Brooks et al., 2016 ; Trivedi et al., 2019 ) and biocomplexed selenium ( Ahmad et al., 2012 ; Göçmen et al., 2016 ; Safdari-Rostamabad et al., 2017 ; Li et al., 2018 ; Silva et al., 2019 ) in broiler diets on weight gain, feed intake, and feed conversion, regardless of mineral source and inclusion levels.

With advancing age, males showed higher values of average weight, weight gain, and feed intake. Variations in performance between birds of different sex may be related to differences in growth rates. According to Mitrovic et al. (2011) and Sarica et al. (2014) , males have higher growth potential and higher weight gain rates, due to higher production of anabolic hormones, especially androgens, which stimulate growth and muscle mass gain in animals. Thus, males have higher feed intake, which may be due to their higher demand for nutrients to express their higher growth potential ( Santos et al., 2005 ). The results of the present study are in agreement with those of Santos et al. (2005) , Faria et al. (2011) , and Del Castilho et al. (2013) , who observed higher average weight gain for males, and with Santos et al. (2005) and Dourado et al. (2009) , in which males presented higher values of feed intake.

Regarding the feed conversion index, males presented better values throughout the production period (1 to 90 days), which indicates that they are more efficient in converting the consumed feed into weight gain. This result is in agreement with those of Santos et al. (2005) , Faria et al. (2011) , and Del Castilho et al. (2013) , who observed better indices of feed conversion for males.

In general, the supplementation with biocomplexed minerals in broiler diets may cause changes in tissue metabolism, especially in muscle tissue ( Zia et al., 2017 ). Chromium is an active component of glucose tolerance factor (GTF) and, therefore, is responsible for promoting greater insulin-binding to cell receptors, thereby increasing the sensitivity of target cells and enhancing insulin action ( Huang et al., 2016 ). Selenium, in turn, acts on the activation of the enzymes iodothyronine deiodases, which promote dehydration of the phenolic or tyrosine ring of the thyroxine molecule (T4), increasing the production and circulating levels of T3 ( Ahmad et al., 2012 ).

Thus, the supplementation with both minerals in the diet may cause increased protein synthesis in muscle tissue, which may be responsible for an increase in the main carcass characteristics. However, in the present study, there were no changes in carcass characteristics with biocomplexed chromium and selenium supplementation. Other authors found no influence of the supplementation of biocomplexed chromium ( Souza et al., 2010 ; Zheng et al., 2016 ; Huang et al., 2016 ; Arif et al., 2019 ; Ognik et al., 2020 ) and selenium ( Oliveira et al., 2014 ; Safdari-Rostamabad et al., 2017 ; Silva et al., 2019 ) on the main carcass characteristics, either. The effectiveness of supplementation basically depends on the levels and bioavailability of the source of minerals to be added to the diet, and these factors may be insufficient to promote alterations in tissue metabolism that allow an increase in the main carcass characteristics ( Huang et al., 2016 ; Li et al., 2018 ).

In a great part of the works in the literature, males presented higher values of live weight ( Mitrovic et al., 2011 ; Sarica et al., 2014 ; Cruz et al., 2018 ) and carcass yield ( Del Castilho et al., 2013 ), which is in agreement with the results observed in the present study. This behavior is related to sexual dimorphism, in which males present a higher growth potential, with greater muscle deposition capacity due to the synthesis of androgen hormones, responsible for protein anabolism, in addition to better developed bone structure ( Rondelli et al., 2003 ).

The fact of that females present higher mean of breast yield and males higher thigh yield and drumstick weight may be related to the effect of sexual dimorphism, and this behavior has been observed in other studies ( Dourado et al., 2009 ; Faria et al., 2011 ; Del Castilho et al., 2013 ; Sarica et al., 2014 ; Cruz et al., 2018 ). Males presented greater development of the secondary parts of the carcass, such as wing, back, head plus neck, and feet. This is due to the fact that males have greater muscle tissue growth potential as well as better developed bone structure with greater limb development ( Rondelli et al., 2003 ; Mitrovic et al., 2011 ; Sarica et al., 2014 ). According to Madeira et al. (2006) , males present more intense physical activity, which is responsible for greater development of the leg muscles.

In general, sex differences in relation to the edible viscera weight can be attributed to differences in body proportions between males and females, since males have higher live weight and proportionally higher internal organ weight ( Mitrovic et al., 2011 ). This justifies the fact that, in the present study, males showed higher values of heart, liver, and gizzard weight. However, regarding yield values, females presented the highest means of yield of liver and total edible viscera (heart, liver, and gizzard). This may be due to the lower live weight values found for female birds, since the edible viscera yield was calculated in relation to the values of live weight.

Regarding the weight and yield of abdominal fat, females showed higher means, which are probably due to hormonal differences and growth rates between the sexes. According to Rondelli et al. (2003) , females show earlier maturity and, thus, there is a reduction in muscle tissue growth, concomitantly with an increase in the deposition of fat in the carcass, compared with males at the same age. In addition, female hormones, especially estrogen, increase body fat deposition ( Kessler and Silva, 2017 ). With similar results, Santos et al. (2005) , Faria et al. (2011) , Mitrovic et al. (2011) , Sarica et al. (2014) , and Cruz et al. (2018) found higher values of abdominal fat yield for females.

The inclusion of biocomplexed sources of chromium and selenium in the diet did not cause changes in the average feed cost (AFC) and in the cost per kilogram of weight gain (CWG), considering the entire production period (1 to 90 days). However, from the point of view of the manager, the inclusion of any ingredient that burdens the slightest the dietary costs is interesting as long as it promotes production gains in the stock. The control treatment was more economically efficient; however, its difference in relation to the chromium and selenium treatments, based on the economic efficiency index (EEI) values, was lower than 1%. Regarding sex, males presented higher AFC, lower CWG, and better economic efficiency, based on the EEI values. This indicates that males, despite presenting the highest feed cost during the growing period, are more economically efficient, since the cost to gain one kilogram of live weight is lower when compared with females.

The addition of biocomplexed chromium and selenium in broiler diets does not affect performance, carcass characteristics, and feed costs. Regarding sex, males present better performance and carcass characteristics, in addition to being economically more viable.