Chia (Salvia hispanica) seeds are a plant source known to be richest in omega-3 (ω-3) fatty acids (Coates and Ayerza, 2009; Nitrayová et al., 2014). They are also a good protein source (19 to 27 g/100 g; Weber et al., 1991) and have an excellent amino acid profile (Nitrayová et al., 2014). Moreover, unlike other plant sources of ω-3 such as linseed, chia has no known antinutritional factor for poultry (Azcona et al., 2008) and has antioxidant activity (Taga et al., 1984). However, it has a considerable fiber content (37.5 g/100 g; Marineli et al., 2014), which may influence the nutrient digestibility of poultry feed.

The dietary chia potential to improve the nutritional quality of broiler meat was confirmed by our research group (Mendonça et al., 2020). However, poultry production requires knowledge of energy content of the feedstuffs to provide adequate energy to the birds (Junqueira et al., 2005), in addition to knowledge of the nutrient digestibility and use by poultry to maximize their productive performance. According to Alvarenga et al. (2015), the energy value and dietary nutrient metabolizability can influence broiler performance.

Thus, to formulate more accurate diets, the chemical composition, metabolizability coefficients, and energy values must be constantly determined for new and conventional ingredients such as chia and soybean, respectively. In addition, alternative ingredients to soybean are important, since the soybean price has been increasing and its global supply is limited (Kierończyk et al., 2018).

The energy values and metabolizability coefficients for chia oil and seed were not determined to date, as well as the effects of their use in replacement of soybean on physiological and metabolic parameters of broilers are still little explored (Ayerza et al., 2002; Azcona et al., 2008; Komprda et al., 2013). Therefore, two experiments were carried with broilers from 29 to 42 days of age for the nutritional evaluation of dietary chia. Experiment I was conducted to determine the nitrogen-corrected apparent metabolizable energy (AMEn) and the apparent metabolizability coefficients of gross energy and ether extract of chia seeds and oil, toasted soybean grain (TSG), and soybean oil. The second experiment was carried with the objective of evaluating the effects of dietary feedstuffs on performance, carcass and cut yields, blood parameters, and lipogenic enzyme activity in broilers, besides the apparent metabolizability coefficients of nutrients and the AMEn of the diets.

The experiments with animals were conducted according to the institutional committee on animal use (protocol number 074/16) and were conducted in Lavras, a city located in the south region of the Minas Gerais state, at an altitude of 919 m and geographic coordinates 21°14'43" South latitude and 44°59'59" West longitude.

Chia seeds were purchased from a local market in São Paulo, SP, Brazil, the chia oil in Rio de Janeiro, RJ, Brazil, and TSG in Três Corações, MG, Brazil. Soybean oil and other ingredients were purchased from a local market in Lavras, MG, Brazil. To obtain the TSG, the soybean grains were heated at 170 ℃ for a period of 1.5 min and subsequently brought to a silo remaining for 30 min (where they reached 150 ℃). The grains were then removed and cooled to room temperature and posteriorly grounded (2 mm). Chia seeds were added whole to the feed because, according to Nitrayová et al. (2014), they do not need to be ground to release their nutrients. The same feedstuffs were used in all experiments and their chemical composition and gross energy were determined (Table 1).

Table 1

Item	Soybean oil	Chia oil	TSG2	Chia seeds
Gross energy (MJ/kg as fed)	39.49	39.23	22.30	24.53
Dry matter	998.5	1000.0	929.6	927.8
Crude protein	-	-	323.2	178.3
Ether extract	-	-	185.3	270.4
Ash	-	-	4.24	3.64
Acid detergent fiber	-	-	155.5	347.5
Neutral detergent fiber	-	-	280.4	561.7
Total dietary fiber	-	-	88.1	308.0

1 All data are the results of chemical analysis conducted in triplicate.

One-day-old broilers were purchased from a commercial hatchery and reared in a conventional broiler shed for 28 days, constituting a single batch. On day 29, the broilers were weighed, homogenized by weight range, and distributed into experiment I (metabolism trial) and experiment II (subdivided in metabolism and performance trials). During the pre-experimental (1 to 28 days of age) and experimental (29 to 42 days of age) periods, water and feed were provided ad libitum. The feeds (mashed) were composed based on corn and soybean meal, formulated to meet the broilers’ nutritional requirements by the recommendations of Rostagno et al. (2017) (Tables 2 and 3). In the pre-experimental phase, starter (1 to 7 days) and growing (8 to 28 days) diets were provided to the broilers.

Table 2

Ingredient (g/kg)			Experimental diet (29–42 days)
	Pre-experimental diet		Experiment I1	Experiment II (performance and metabolism trials)2
	1–7 days	8–28 days	Reference diet	Soybean oil	Chia oil	Toasted soybean grain (TSG)	Chia seeds
Corn	553.3	594.8	651.00	651.00	651.00	644.50	644.50
Soybean meal	381.8	340.6	292.50	292.50	292.50	160.00	160.00
Dicalcium phosphate	19.0	14.6	12.00	12.00	12.00	11.60	11.60
Limestone	9.10	9.1	8.00	8.00	8.00	7.70	7.70
Sodium chloride (NaCl)	5.10	4.8	4.50	4.50	4.50	4.50	4.50
DL-Methionine, 99%	3.65	2.7	2.50	2.50	2.50	2.50	2.50
L-Lysine HCL, 78%	3.10	2.2	2.10	2.10	2.10	2.40	2.40
L-Threonine, 98.5%	1.10	0.7	0.40	0.40	0.40	0.50	0.50
L-Valine, 99%	-	-	0.20	0.20	0.20	0.40	0.40
Mineral supplement3	0.50	0.5	0.50	0.50	0.50	0.50	0.50
Vitamin supplement4	0.50	0.4	0.40	0.40	0.40	0.50	0.50
Choline chloride, 60%	0.50	0.45	0.40	0.40	0.40	0.40	0.40
Anticoccidial5	0.50	0.5	0.50	0.50	0.50	0.50	0.50
Soybean oil	21.8	28.6	25.00	25.00	-	-	-
Chia oil	-	-	-	-	25.00	-	-
TSG	-	-	-	-	-	164.00	-
Chia seed	-	-	-	-	-	-	164.00

1 Soybean and chia oils replaced reference diet in 100 g/kg; TSG and chia seeds replaced reference diet in 250 g/kg.

Table 3

Item			Experimental diet (29–42 days)
	Pre-experimental diet		Experiment I1	Experiment II (performance and metabolism trials)2
	1–7 days	8–28 days	Reference diet	Soybean oil	Chia oil	Toasted soybean grain (TSG)	Chia seeds
Metabolizable energy (MJ/kg)3	12.37	12.78	12.99	12.99	-	12.99	-
Metabolizable energy (MJ/kg)4	-	-	13.00	13.00	13.11	13.33	12.72
Crude protein3,5	217.00	201.40	187.50	187.50	187.50	187.30	163.20
Crude protein6	-	-	179.90	179.90	179.90	180.70	155.20
Digestible methionine + cystine3	9.50	8.20	7.70	7.70	7.70	7.60	-
Digestible lysine3	13.30	11.60	10.40	10.40	10.40	10.40	-
Digestible threonine3	8.50	7.60	6.80	6.80	6.80	6.80	-
Digestible valine3	-	-	8.20	8.20	8.20	8.10	-
Total phosphorus3,5	6.90	6.00	6.90	6.90	6.90	6.80	6.10
Available phosphorus3,5	4.70	3.80	3.30	3.30	3.30	3.20	3.40
Calcium3,5	9.30	8.10	6.90	6.90	6.90	6.80	7.20
Sodium3,5	2.20	0.20	2.00	2.00	2.00	2.00	1.90
Acid detergent fiber6	-	-	5.23	5.23	5.30	7.01	10.16
Neutral detergent fiber6	-	-	17.12	17.12	17.03	19.96	23.81
Total dietary fiber6	28.1	26.8	26.80	26.80	26.80	34.20	51.00

1 Soybean and chia oils replaced reference diet in 100 g/kg; TSG and chia seeds replaced reference diet in 250 g/kg.

Experiment II was subdivided into two trials in which the same feedstuffs used in experiment I (Table 1) were evaluated. A total of 192 male broilers in the period from 29 to 42 days of age were used in experiment II, of which 120 broilers were used in the performance trial and 72 in the metabolism trial. In the performance trial, broiler performance, carcass and cut yields, serum lipid profile, and lipogenic enzyme activity were evaluated. The metabolism trial was carried simultaneously with the performance trial, but in a metabolic room. In the metabolism trial, the AMEn, AMCEE, AMCGE, and apparent metabolizability coefficient of dry matter (AMCDM) and crude protein (AMCCP) of the experimental diets were determined.

In experiment II, TSG was totally replaced by chia seeds (164 g/kg) because of the lack of prior knowledge about the AMEn value and utilization of chia nutrients by broilers. Thus, these values and values of chemical composition, of chia, which were determined in Experiment I, enabled determining later that the nutritional levels of the diet with chia seeds in experiment II were below the nutritional requirements recommended by both the NRC (1994) and Brazilian Tables (Rostagno et al., 2017).

The NRC (1994) and Rostagno et al. (2017) recommend, respectively, 14.08 and 13.29 MJ/kg of AMEn for broilers in the rearing period evaluated in the present study. However, the final AMEn of the diet containing chia seeds was 12.72 MJ/kg (Table 3). A similar condition was observed in relation to the CP content of the diet containing chia seeds, because in this diet, the final CP was 15.52%, while the values recommended by the NRC (1994) and Rostagno et al. (2017) are of 20.0 and 19.5%, respectively.

Kamran et al. (2008) observed a linear increase of FI when broilers received diets with reduced contents of CP and metabolizable energy, in all rearing periods. Moreover, according to Leeson et al. (1993), FI is regulated by the dietary energy value. Thus, the lowest metabolizable energy and CP values of the diet containing chia seeds in experiment II may explain the higher FI by the broilers that received this experimental diet, possibly as a mechanism to meet their nutritional requirements. On the other hand, the broilers fed diet containing soybean oil during the experimental period exhibited higher FI, possibly because unlike of the other experimental dietary treatments, they did not have to adapt to the feed, since the diet provided in the pre-experimental phase (from 1 to 28 days of age) also contained soybean oil in its formulation.

According to Alvarenga et al. (2015), dietary energy influences the development and performance of broilers. Therefore, although broilers that received chia seeds in their diet consumed more feed, they did not exhibit higher WG, possibly because they used chia seeds were not ground, reflecting in lower AMCEE, AMCGE, and AMEn values (Table 8). Chia seeds were added whole to the feed because, according to Nitrayová et al. (2014) they do not need to be ground to release their nutrients. However, in the present study, the results suggest that the nutrient use by the broilers could have been maximized if the chia seeds had been ground prior to its inclusion into the feed.

Moreover, chia seeds also have high fiber content. Marineli et al. (2014) determined 37.5% of dietary fiber in the chia seeds. In the present study, the chia seeds exhibited 30.8% of total dietary fiber (Table 1). According to Janssen and Carré (1989), increased dietary fiber content may reduce nutrient and energy use. In the case of chia, this may occur because when the seeds get in contact with water, they gelatinize to form a thick mucilage layer, which acts as a physical barrier changing the passage rate through the gastrointestinal tract and hindering the action of digestive enzymes (Bedford, 2000; Alzueta et al., 2003).

The WG and FCR of the broilers that received dietary chia seeds corroborate the results of Ayerza et al. (2002), who also observed lower WG and FCR in broilers fed diet containing chia. Overall, it was observed that the inclusion of soybean and chia oils in the diet resulted in higher WG than the inclusion of TSG and chia seeds, probably because the AMEn, AMCEE, and AMCGE of the oils were higher than the values determined for TSG and chia seeds (Table 4). Moreover, current broiler lines have high energy requirements; thus, oil must be used in their diets.

Similar to the results of the present study in which the experimental diets did not change the carcass yield, Azcona et al. (2008) also did not observe statistical difference in the carcass yield when the broiler diets were formulated with linseed, canola, chia seeds, or chia flour. However, Azcona et al. (2008) observed that broilers fed diets containing linseed and chia seeds exhibited lower abdominal fat deposition than did those that received the control diet. These authors attributed this difference to changes in the β-oxidation rate of polyunsaturated fatty acids (PUFA) between the evaluated diets, but in the present study, there were no effects of the experimental diets on abdominal fat deposition.

The AMEn values determined for the soybean oil and TSG in experiment I (37.35 and 15.85 MJ/kg DM, respectively) were higher than the values reported by Rostagno et al. (2017) (36.86 MJ/kg DM for soybean oil and 14.59 MJ/kg DM for TSG). Moreover, the AMEn of the TSG determined in this work was also higher than the 13.82 MJ/kg DM cited by the NRC (1994). This underscores the importance of constant updating of the feed table values to contribute with dietary formulations increasingly accurate and efficient.

Although the gross energy of chia oil was 0.26 MJ/kg DM lower than that in the soybean oil (Table 1), the AMEn determined for chia oil was 0.14 MJ/kg higher than the value determined for soybean oil (Table 4), demonstrating that chia oil is highly used by broilers. In its turn, the AMEn of the TSG for broilers was higher than determined for chia seeds (Table 4). However, unlike chia seeds, TSG was initially ground and then was included in the feed.

According to Kris-Etherton et al. (1988), diets rich in PUFA reduce the cholesterol levels. However, the broilers fed diet containing chia seeds exhibited higher serum total cholesterol level. Besides rich in ω-3 series fatty acids, chia also has considerable amounts of protein and other nutrients (Nitrayová et al., 2014). However, in the present study, the results suggest that chia seeds were not fully digested because they were incorporated whole into the feed. Thus, as a consequence, a variation may have occurred in the quantity and profile of circulating nutrients during the absorptive period, which may have resulted in intensified liver cholesterol biosynthesis. Similarly, in a study conducted with laying hens, Ayerza and Coates (1999) verified that the cholesterol content in the yolk was higher when the diet was formulated with chia seeds. However, further studies are needed to elucidate these results.

The use of chia seeds and oil in broiler diets did not reduce the HDL-cholesterol and triglycerides levels, contrary to what was observed by Ayerza and Coates (2005, 2007) in rats. However, the lipoprotein metabolism in rats is not identical to that observed in broilers, justifying such differences. The readily available PUFA present in chia and soybean oils provided the lowest LDL cholesterol levels in the birds. Weintraub et al. (1988) evaluated the inclusion of saturated fatty acids, ω-6 PUFA, and ω-3 PUFA in human diets and found that both types of PUFA decreased the total cholesterol, triglycerides, VLDL-cholesterol, and LDL cholesterol levels. According to these authors, the ω-6 and ω-3 PUFA present in chylomicrons are more susceptible to lipoprotein lipase action, which contributed to the reduced plasma lipoproteins levels.

Shimomura et al. (1990) reported that the diets rich in PUFA increased the lipoprotein lipase activity in rats. In addition, Beynen and Katan (1985) found that the liver preferentially converts PUFA to ketone bodies rather than to VLDL. In plasma, VLDL is rapidly converted to LDL, which is the main carrier of cholesterol. Thus, with lower VLDL formation, less LDL will be circulating in the plasma. These authors also reported that an increase in sterol excretion occurs in individuals who consume more PUFA, which may have contributed to the lower LDL cholesterol levels determined in the broilers that received soybean or chia oils in the diet.

In birds, lipogenesis occurs mainly in the liver, unlike in pigs where lipid synthesis is highest in adipose tissue (O’Hea and Leveille, 1968; Leveille et al., 1975). Glucose-6-phosphate dehydrogenase and malic enzymes provide reducing equivalents in the form of NADPH for de novo lipogenesis in the liver. Therefore, these enzymes can be used as key indicators of lipid biosynthesis (Alvarez et al., 2000). In the present study, only the malic enzyme activity was influenced by diet, in which the highest activity was determined when the broilers received the dietary chia seeds. The activity of the malic enzyme was higher than glucose-6-phosphate dehydrogenase, corroborating O’Hea and Leveille (1968), who reported that the malic enzyme is the primary source of reducing equivalents (NADPH) for lipid synthesis in the liver of birds.

The diet formulated with chia seeds increased the activity of the malic enzyme probably due to its lowest CP level content, because Adams and Davis (2001) and Rosebrough et al. (2008) concluded that the reduced CP content in the diet increase the mRNA expression of the malic enzyme, which can explain the increased activity of malic enzyme in the liver extract of broilers.

These observations may help explain the results obtained in this study since chia seeds were not fully digested by the broilers. Thus, part of the nutrients of the seeds, such as protein, may not have been fully digested, leading to a greater concentration, digestion, and absorption of carbohydrate in relation to the other nutrients in the diet. Moreover, according to Stabile et al. (1998), dietary PUFA inhibit lipogenic enzyme expression in rats, which is consistent with the lower values determined for malic enzyme in broilers fed diets containing soybean oil, chia oil, and TSG, in which nutrient use by broilers was generally higher.

Therefore, the energy values determined for soybean oil, chia oil, TSG, and chia seeds were 37.35, 37.49, 15.85, and 8.43 MJ/kg DM, respectively. The metabolizability of gross energy and ether extract coefficients ranged from 0.32 to 0.96 and from 0.54 to 0.93, respectively, and the oils presented the highest coefficients.

Chia oil presents a similar AMEn to soybean oil as well as good metabolizability coefficients of ether extract and gross energy, without compromising the performance and carcass characteristics of broilers.