Effects of dietary starch-to-lipid ratios and amino acid levels on broilers exposed to cyclic heat stress during the finisher period

Abstract

This study aimed to assess the effects and interactions between amino acid (AA) content and starch-to-lipid ratios (S:L) in finisher diets on Ross 308 broilers exposed to cyclic heat stress. Ross 308 mixed sex broilers (21 d of age) were fed isocaloric finisher diets formulated to have 80, 90, 100, 110, 120, or 130% of their estimated AA requirements (AAR) and S:L ratios of either 4:1 or 10:1, resulting in a total of 12 dietary treatments arranged in a 6 × 2 factorial design. Birds were housed at a final estimated stocking density of 31 kg/m² in 8 rooms, each containing 12 pens. Rooms were maintained at a temperature of 21°C from d 21 to 27. Starting on d 28, birds were subjected to cyclic heat stress at 31°C for 12 h/d, followed by a cooler temperature of 21°C, with a minimum humidity of 50%. Body weight and residual feed weights were recorded on d 21, 28, and 35, which were used to calculate BW gain (BWG) and feed-to-gain ratios (F:G). Birds fed 80, 100, and 120% AAR were slaughtered on d 35 to determine meat yield. Starch-to-lipid ratios did not affect BWG or F:G ratios from d 21 to 35. Based on the linear (LBL) and quadratic (QBL) broken line models, the estimated breakpoints from d 21 to 35 were 110% and 122% AAR for BWG of 1.474 and 1.475 kg, respectively; whereas estimated breakpoints for LBL and QBL were 114% and 129% AAR respectively, for F:G ratios of 1.525 and 1.515. There was no effect of the dietary treatments on oxidative stress biomarkers (d 31). Highest breast meat yield (% live weight) was observed in birds fed 120% AA. In conclusion, while differing S:L ratios did not affect BWG or F:G ratios, finisher diet (d 21-35) AAR in the range of 114 to 129% improved performance of a population of broilers when exposed to heat stress from d 28-35.

Introduction

Heat stress poses a significant challenge to the poultry industry, requiring cost-effective solutions to mitigate its adverse effects on animal health and performance. Birds primarily rely on panting for thermoregulation, which becomes less effective in high temperature and humidity environments (Nawab et al., 2018). Broilers, in particular, are highly susceptible to heat stress due to their rapid growth and associated metabolic demands (Awad et al., 2020; Goo et al., 2019). To reduce metabolic heat production, heat stressed birds typically reduce their activity levels and feed intake, leading to lower growth and poorer feed efficiency (Teyssier et al., 2022; Zaboli et al., 2018). Heat stressed broilers will also exhibit reduced meat yield, increased fat deposition, and reduced meat quality characteristics (Ghazalah et al., 2008; Teyssier et al., 2022; Zaboli et al., 2018). Heat stress also increases production of reactive oxidative species, overwhelming antioxidants such as catalase and glutathione, leading to increased oxidation of protein and lipids producing compounds such as thiobarbituric reactive substances (TBARS) and protein carbonyl (Habashy et al., 2019).

Nutritional strategies and feeding practices may help mitigate effects of the decreased feed intake in broilers resulting from heat stress. Dietary proteins produce more heat per unit of calorific value compared to carbohydrates (Black, 1995) and reducing the heat increment by lowering CP levels and supplementing with amino acids (AA) has been proposed as a nutritional approach to alleviate the adverse effects of heat stress (Awad et al., 2019). Conversely, low feed intake in heat-stressed birds, particularly those on low-protein diets, may lead to AA deficiencies; therefore, providing an excess of essential AA may help mitigate these nutrient deficiencies during periods of heat stress (Furlan et al., 2004). Recent findings by Maharjan et al. (2020) support the idea that increasing the levels of digestible AA to 110 to 120% of estimated AA requirements (AAR) can improve performance of broilers, particularly in hot environmental conditions.

Another dietary recommendation is to increase the proportion of fat in the diet compared to starch, as fat produces a lower heat increment at the same level of ME intake. When dietary fat is utilized for energy, it generates less heat per unit of energy consumed compared to starch or protein (Syafwan et al., 2012). Consequently, a diet high in fat results in lower heat production compared to a diet with the same ME derived from starch or protein (Balnave and Brake, 2005). Therefore, adjusting the ratio of dietary starch to lipid may serve as an effective strategy to mitigate the adverse effects of heat stress.

In a previous study (Duhra et al., 2025), a simplex mixture model was used to determine an optimum diet composition for broilers undergoing cyclic heat stress during the finisher phase. The optimal diet had starch-to-lipid (S:L) ratio of 4:1, an AMEn of 3,089 kcal/kg and a digestible AA level of 1.01% digestible lysine as fed (Duhra et al., 2025). Broiler performance was reduced in birds fed increasing levels of the diet with a S:L of 20:1 relative to the animals fed the other basal diets with a S:L of 4:1 under cyclic heat stress. However, it is important to note that unlike factorial models, the component proportions in mixture experiments cannot vary independently since they are constrained to sum to 100% (Bondari, 2005) thus, it was only possible to compare the effects of a diet relative to the other diets. This limitation did not allow the direct investigation of potential effects of S:L ratio, dietary AA density, and AME. In this context, the current trial aimed to further our understanding. Specifically, we wanted to assess how six graded AA levels and two S:L ratios arranged as a 6 × 2 factorial affect growth performance, meat yield, biomarkers for oxidative damage, and nutrient retention of Ross 308 broilers exposed to cyclic heat stress (d 28-35) during the finisher period (d 21-35).

Materials and methods

The protocol for this trial was approved by the University of Saskatchewan’s Animal Care Committee (AUP 20210085) and was performed in accordance with the recommendations outlined in the Guide on the Care and Use of Experimental Animals by the Canadian Council on Animal Care (2009).

Diets

Broilers were fed common commercial starter (0-10 d) and grower diets (10-21 d) formulated according to AMINOChick ® 3.0 software (Table 1; Lemme, 2021). Samples of corn, wheat middlings, soybean meal, and corn gluten meal were analyzed using AMINONir Advanced program (Evonik Operations GmbH, Hanau, Germany) for proximate analyses of starch, lipid, energy, CP, and digestible AA content. On d 21, broilers were transitioned to the experimental finisher diets (Table 2) formulated to contain 3,089 kcal/kg AMEn and six levels of digestible Lys (dLys) corresponding to 80%, 90%, 100%, 110%, 120%, and 130% of the estimated requirement of 1.01% dLys (Duhra et al., 2025) with essential AA formulated to a fixed ratio to dLys derived from AMINOChick ® 3.0 software (Lemme, 2021). Consequently, dietary CP increased with the increase in the balanced AA content. The diets were also formulated to contain S:L ratios of 4:1 or 10:1 in a factorial arrangement for 12 total diets based on analyzed values. Celite (89% diatomaceous earth, EP Minerals, Reno, NV) was included in the finisher diets at 0.5%, as fed, to provide a marker to calculate to allow calculation of nutrient retention.

Table 1.

Starter and grower diet formulations (% as fed).

Ingredient (%)	Starter diet (0-10 d of age)	Grower diet (10-21 d of age)
Corn	50.879	57.014
Soybean meal, 45% CP	37.955	32.601
Canola oil	3.442	4.173
Corn gluten meal, 60% CP	2.889	1.816
Dicalcium phosphate	2.127	1.921
Limestone	0.861	0.791
Salt	0.319	0.318
Sodium bicarbonate	0.100	0.107
Poultry premix1	0.500	0.500
Choline chloride 60%	0.101	0.103
Biolys 602	0.350	0.280
DL-methionine	0.326	0.275
L-threonine	0.085	0.064
L-valine	0.066	0.037
Estimated Composition % as fed
AMEn	3,000	3,100
CP	23.686	20.994
Calcium	0.960	0.870
Available phosphorus	0.480	0.435
Digestible Lys	1.280	1.110
Digestible Met + Cys	0.930	0.820
Digestible Met	0.641	0.557
Digestible Thr	0.810	0.710

¹Premix contains 2,400,000 IU/kg vitamin A, 700,000 IU/kg vitamin D3, 20,000 IU/kg vitamin E, 4,000 µg/kg vitamin B12, 50,000 µg/kg biotin, 600 mg/kg menadione, 500 mg/kg thiamine, 1,400 mg/kg riboflavin, 6,000 mg/kg pantothenic acid, 1,000 mg/kg pyridoxine, 14,000 mg/kg niacin, 400 mg/kg folic acid, 12,000 mg/kg iron, 4,000 mg/kg copper, 24,000 mg/kg manganese, 22,000 mg/kg zinc, 500 mg/kg iodine, and 60 mg/kg selenium (DSM Nutritional Products Canada Inc. Ayr, Ontario, Canada).

²Minimum 60% l-lysine (Evonik Industries GmbH, Hanau-Wolfgang, Essen, Germany).

Table 2.

Finisher diet formulations and analyzed composition (% as fed; d21-35).

Amino acid density	80%	90%	100%	110%	120%	130%	80%	90%	100%	110%	120%	130%
Starch-to-lipid ratio	4:1	4:1	4:1	4:1	4:1	4:1	10:1	10:1	10:1	10:1	10:1	10:1
Corn	47.56	47.83	47.62	46.88	47.09	46.93	75.05	72.83	70.59	68.19	65.81	63.55
Wheat middlings	31.09	26.18	22.23	19.05	14.32	10.17	0.00	0.00	0.00	0.00	0.00	0.00
Soybean meal 45% CP	5.58	11.04	15.25	18.71	23.97	28.41	15.21	15.30	15.52	15.83	16.32	16.58
Corn gluten meal 60% CP	2.88	2.30	2.42	2.98	2.51	2.52	1.99	3.76	5.50	7.20	8.78	10.49
Canola oil	6.52	6.34	6.13	5.93	5.74	5.54	1.91	1.77	1.63	1.54	1.45	1.32
Monocalcium phosphate	1.45	1.43	1.42	1.40	1.34	1.37	1.48	1.48	1.47	1.50	1.47	1.46
Limestone	1.30	1.28	1.26	1.25	1.24	1.21	1.27	1.28	1.27	1.27	1.28	1.27
Poultry premix1	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50
Salt	0.27	0.28	0.28	0.29	0.29	0.29	0.29	0.30	0.29	0.30	0.30	0.29
Sodium bicarbonate	0.30	0.29	0.28	0.28	0.27	0.26	0.27	0.26	0.25	0.25	0.24	0.24
Potassium carbonate	0.40	0.32	0.25	0.20	0.12	0.05	0.52	0.53	0.54	0.54	0.53	0.53
Choline chloride 60%	0.16	0.14	0.12	0.11	0.09	0.07	0.15	0.15	0.15	0.15	0.14	0.14
Biolys 602	0.62	0.61	0.64	0.68	0.67	0.69	0.41	0.56	0.70	0.84	0.97	1.12
DL-Methionine	0.20	0.25	0.30	0.34	0.40	0.45	0.19	0.23	0.26	0.30	0.35	0.39
L-Arginine	0.19	0.18	0.19	0.22	0.21	0.22	0.11	0.19	0.26	0.34	0.41	0.48
L-Threonine	0.15	0.16	0.18	0.20	0.21	0.23	0.08	0.12	0.16	0.20	0.23	0.27
L-Isoleucine	0.14	0.14	0.15	0.17	0.17	0.18	0.05	0.08	0.12	0.16	0.19	0.23
L-Valine	0.11	0.13	0.16	0.18	0.20	0.22	0.04	0.09	0.13	0.17	0.22	0.26
Glycine	0.08	0.09	0.12	0.15	0.17	0.19	0.00	0.07	0.13	0.19	0.25	0.31
L-Tryptophan	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.01	0.02	0.04	0.05	0.06
Celite3	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50	0.50
Estimated composition
AMEn	3,089	3,089	3,089	3,089	3,089	3,089	3,089	3,089	3,089	3,089	3,089	3,089
CP	16.0	17.5	18.6	19.6	21.1	22.0	15.9	16.8	18.4	19.3	20.8	21.9
Starch	38.0	37.0	35.9	34.8	33.7	32.7	47.7	46.7	45.6	44.4	43.1	40.2
Ether extract	9.5	9.2	9.0	8.7	8.4	8.2	4.8	4.7	4.6	4.4	4.3	4.2
Calcium	0.89	0.89	0.89	0.89	0.89	0.89	0.89	0.89	0.89	0.89	0.89	0.89
Available phosphorus	0.44	0.44	0.44	0.44	0.44	0.44	0.44	0.44	0.44	0.44	0.44	0.44
Digestible Lys	0.81	0.91	1.01	1.11	1.21	1.31	0.81	0.91	1.01	1.11	1.21	1.31
Digestible Met+Cys	0.62	0.69	0.77	0.85	0.92	1.00	0.62	0.69	0.77	0.85	0.92	1.00
Digestible Met	0.40	0.44	0.51	0.56	0.62	0.69	0.41	0.45	0.50	0.55	0.60	0.65
Digestible Thr	0.54	0.60	0.67	0.74	0.80	0.87	0.54	0.60	0.67	0.74	0.80	0.87
Analyzed diet composition (% as fed)
Crude protein	15.65	17.25	18.12	20.28	21.08	21.71	15.65	16.82	18.68	19.42	20.77	22.61
Starch	38.6	40.1	36.7	36.2	35.1	35.7	47.1	49.1	45.5	45.6	43.5	44
Ether extract	8.9	8.2	8.7	8.6	8.7	8.2	6.0	4.9	5.8	4.8	5.0	4.0
Met	0.434	0.498	0.563	0.635	0.710	0.782	0.449	0.517	0.563	0.626	0.681	0.775
Cys	0.278	0.292	0.303	0.329	0.338	0.340	0.273	0.290	0.312	0.320	0.334	0.360
Met+Cys	0.712	0.790	0.866	0.964	1.048	1.122	0.722	0.807	0.875	0.946	1.015	1.135
Lys	0.928	1.035	1.126	1.250	1.296	1.427	0.843	0.984	1.058	1.117	1.230	1.307
Thr	0.630	0.707	0.780	0.892	0.915	0.963	0.623	0.698	0.747	0.800	0.891	0.944
Arg	0.981	1.105	1.207	1.374	1.385	1.481	0.944	1.065	1.151	1.211	1.304	1.450
Ile	0.646	0.708	0.785	0.888	0.929	0.978	0.657	0.717	0.788	0.814	0.881	0.965
Leu	1.422	1.484	1.563	1.755	1.765	1.833	1.633	1.750	1.951	2.052	2.154	2.322
Val	0.786	0.863	0.956	1.073	1.120	1.181	0.768	0.858	0.945	0.970	1.072	1.166
His	0.365	0.399	0.434	0.477	0.491	0.517	0.390	0.412	0.432	0.429	0.453	0.474
Phe	0.712	0.774	0.840	0.944	0.957	1.006	0.785	0.83	0.900	0.919	0.97	1.045
Gly	0.688	0.763	0.823	0.951	0.957	1.034	0.599	0.692	0.792	0.835	0.907	1.031
Ser	0.685	0.768	0.816	0.933	0.935	0.993	0.752	0.795	0.863	0.881	0.923	0.990
Pro	1.096	1.135	1.189	1.286	1.290	1.319	1.151	1.243	1.375	1.350	1.411	1.498
Ala	0.887	0.920	0.955	1.066	1.056	1.095	0.963	1.034	1.144	1.185	1.240	1.333
Asp	1.092	1.300	1.457	1.674	1.748	1.886	1.300	1.357	1.420	1.411	1.499	1.576
Glu	2.819	3.010	3.205	3.529	3.547	3.682	2.817	2.978	3.233	3.366	3.556	3.808

¹Premix contains 2,400,000 IU/kg vitamin A, 700,000 IU/kg vitamin D3, 20,000 IU/kg vitamin E, 4,000 µg/kg vitamin B12, 50,000 µg/kg biotin, 600 mg/kg menadione, 500 mg/kg thiamine, 1,400 mg/kg riboflavin, 6,000 mg/kg pantothenic acid, 1,000 mg/kg pyridoxine, 14,000 mg/kg niacin, 400 mg/kg folic acid, 12,000 mg/kg iron, 4,000 mg/kg copper, 24,000 mg/kg manganese, 22,000 mg/kg zinc, 500 mg/kg iodine, and 60 mg/kg selenium (DSM Nutritional Products Canada Inc. Ayr, Ontario, Canada).

²Minimum 60% l-lysine (Evonik Industries GmbH, Hanau-Wolfgang, Essen, Germany).

³Minimum 89% diatomaceous earth (EP Minerals, Reno, NV).

Housing

This experiment consisted of two trials, where the broilers originated from the same flock. A total of 6,720 mixed sex Ross 308 broiler chicks were housed in eight independently temperature-controlled rooms, each containing 12 pens (2 × 2.3 m, 4.6 m²), with 70 birds per pen. The photoperiod started at 22L:2D, changing to 20L:4D by d 3 (1 h increase in dark per d) with a 15-minute dawn to dusk period. Light intensity was set at 40 lux on d 1 and was gradually reduced to 10 lux by d 9.

Temperatures were initially maintained at 32°C, decreasing to 21°C by d 25. During the heat stress period (d 28-33), the temperature was maintained at 31°C from 8am-8pm (humidex calculated as described by Diaconescu et al., 2022) of 41°C, during the photoperiod) and reduced to 21⁰C for the remaining 10 h (humidex 23°C). A 1 h heat up and cool down period was implemented before and after the heat stress phase. Humidity was between 50 and 65% for the duration of the trial.

Trial one. At 21 d of age, broilers were transitioned to the experimental finisher diets, and the number of birds per pen was reduced to 60, achieving a density of 31 kg/m² based on a final estimated BW at 35 d (Aviagen, 2022). Each room was treated as a block, with all twelve treatments represented once per block, resulting in a total of 8 replicates per treatment. Wood shavings were used as the litter material at a depth of 10 cm. Each pen was equipped with a tube feeder (36 cm pan diameter from d 0-21 and 43 cm diameter until d 22-35) and 6 nipple drinkers.

Trial two. A total of 192 birds, each 21-d-old, were assigned to 48 bioassay cages with 4 cages of 4 birds per treatment group. The birds used in this trial were selected from those removed on d 21 of trial one, where two birds per pen were removed and randomly assigned to bioassay cages for trial two. Each cage housed four birds (dimensions: 46 cm high × 51 cm wide × 51 cm long). Each cage was equipped with two nipple drinkers at the back and a feed trough at the front. Birds in these cages were exposed to heat stress as previously described, with lighting, humidity, and temperature maintained at similar environmental conditions as those in trial one pens. This trial was conducted in a separate room from the rooms used for trial one.

Growth performance

Average BW was measured on d 0, 21, 28, and 35. Measurements were taken on a per pen basis, which allowed for calculation of average BW and BW gain (BWG). Feed refusals were weighed on d 21, 28, 31, and 35, also on a per pen (or cage) basis, allowing the calculation of feed intake per bird for each period. Feed efficiency (feed-to-gain (F:G) ratio) was calculated using the change in BW and feed intake for each period. All culls and mortalities were weighed, and starting from d 7, the cause of mortality was determined by a trained pathologist at Prairie Diagnostic Services Inc. (Saskatoon, Canada).

Carcass yield

Meat yield was measured at the end of the trial from 16 randomly selected birds of each sex from the treatment groups receiving diets with 80%, 100%, and 120% AAR from both S:L ratios (2 males and 2 females per pen; Trial 1). Live weights of the birds were recorded prior to feed withdrawal for a minimum of 8 h. Birds were shackled, slaughtered via stunning with an electric knife (VS200, Midwest Processing Systems, Eden Prairie, MN) at a power level of 5 (circa 0.16 amps, 60 Hz AC), and exsanguinated by cutting the carotid arteries and jugular veins. Carcasses were scalded at 62 to 64°C for 30 s, plucked mechanically, eviscerated, prepared (neck, feet, and uropygial gland removed), chilled in an ice bath for 1 h, and then placed on ice for a minimum of 16 h. Data collected included live weight, carcass weight, Pectoralis (P.) majors and minors, wings, whole thighs, whole drumsticks, and back and rack weights.

Meat quality

Meat quality was assessed on one bird/pen/sex (Trial 1) using the left and right P. major muscles. Whole breast samples were collected during meat yield. Drip loss was measured in the left P. major by blotting with a paper towel, weighing the sample, refrigerating for 24 h, and then blotting dry and weighing. The final weight was divided by the initial weight to determine the percent weight reduction. The right P. major was blotted with a paper towel, weighed, and placed into a Ziploc bag, then placed in a freezer at –20°C. The frozen P. majors were then thawed in a refrigerator for 24 h, blotted dry, weighed, placed into new Ziploc bags and cooked in a hot water bath at 80°C until an internal temperature of 75°C was reached. Samples were cooled to 40°C, removed from the Ziploc bag, blotted dry with a paper towel, and weighed. The thawed and cooked weights were divided by the initial weights to determine percent freeze thaw loss and cook loss.

The pH of the left breast sample was collected at time of slaughter after evisceration and before chilling, and 24 h later after being removed from the carcass to determine pH change using a pH meter in the thickest part of the breast (Accumet, Fisher Scientific, Ottawa, ON, Canada).

Color was recorded at time of final weighing for drip loss. The breast was fileted open and allowed to bloom for 30 minutes at room temperature (22°C). A Minolta color meter (CR-400, Konica Minolta Sensing Americas, Ramsey, NJ) was used to measure meat color at 3 locations on the breast with each location recorded twice with the second scan taken after the meter was rotated 90° to account for differences in light reflection. Readings were averaged for the whole breast.

Physiological biomarkers and bursa size

One bird/pen/sex (Trial 1) was weighed and then euthanized by cervical dislocation, during heat stress on d 31. One 5 g sample each of the P. major and liver tissue was collected, snap frozen in liquid nitrogen (N), and then stored at –80°C. The biomarkers assessed were thiobarbituric acid reactive substances (TBARS; oxidative damage), catalase, and glutathione reductase (antioxidant capacity). Bursa weights were collected for all birds sampled for biomarkers (one bird/sex/pen). Bursa weight for each bird was divided by its’ BW to determine bursa weight as a percent of live weight.

Thiobarbituric acid reactive substances. Approximately 50 mg of P. major or 25 mg of liver tissue were prepared according to manufacturer recommendations (Cayman Chemical Company, Ann Arbor, MI, item no. 10009055) by homogenizing in 250 μL of radioimmunoprecipitation assay (RIPA) buffer, containing 1 mM EDTA. The samples were then centrifuged at 1,600 × g at 4°C for 10 min. Supernatants were collected and prepared using a commercially available kit (µM, Cayman Chemical Company, Ann Arbor, MI, item no. 10009055) and analyzed on a microplate reader at 540 nm (Biotek Epoch 2 Microplate Spectrophotometer (Santa Clara, CA).

Catalase. Approximately 200 mg of the P. major muscle or 100 mg of liver tissue were homogenized in 1 mL of cold 50 mM potassium phosphate buffer, pH 7, containing 1 mM EDTA, in accordance with the manufacturer recommendations (Cayman Chemical Company, Ann Arbor, MI, item no. 707002). After centrifugation at 10,000 × g for 15 min at 4°C, supernatants were prepared using a kit for catalase (nmol/min/ml; Cayman Chemical Company, Ann Arbor, MI, item no. 707002) and analyzed using a microplate reader at 540 nm (Biotek Epoch 2 Microplate Spectrophotometer (Santa Clara, CA).

Glutathione reductase. Approximately 200 g of the P. major muscle or 100 mg of liver tissue were homogenized in 1 mL of cold 50 mM potassium phosphate buffer, pH 7, containing 1 mM EDTA, following manufacturer recommendations (Cayman Chemical Company, Ann Arbor, MI, item no. 703202). After centrifugation at 10,000 × g for 15 min at 4°C, supernatants were prepared using a kit for glutathione reductase (nmol/min/ml; Cayman Chemical Company, Ann Arbor, MI, item no. 703202) and analyzed using a microplate reader at 340 nm (Biotek Epoch 2 Microplate Spectrophotometer (Santa Clara, CA).

Nutrient retention

Excreta were collected from d 32 to 34 from Trial 2. Clean aluminum trays were placed under the cages to collect samples of excreta, which were subsampled and mixed. Excreta were stored at −20°C before being dried at 55°C for 96 h and ground through a 1 mm screen. Dry matter was assessed by drying samples in a 135°C oven for 2 h (AOAC, 2016). Nitrogen (N) was measured using a Leco FP-528 auto-analyzer (St. Joseph, MI), which combusted samples with pure oxygen at 850°C. An aliquot of gas from the combusted sample was then passed through a copper catalyst, which converted nitrous oxides to N₂, with thermal conductivity being used to determine N content (AOAC, 2016). Gross energy was determined using adiabatic bomb calorimetry (Parr 6400 calorimeter, Parr instrument Co. Moline, IL). Acid-insoluble-ash was measured using a procedure from Vogtmann et al. (1975). Nitrogen retention, DM retention, and AMEn were calculated using the following equations:

$$\%Nretention=100*(1-\left(\%NExcreta/\%AIAFeed\right)*\left(\%AIAExcreta/\%NFeed\right)$$

$$\%DMretention=100*(1-\left(\%DMExcreta/\%AIAFeed\right)*\left(\%AIAExcreta/\%DMFeed\right)$$

AMEn = GEFeed – ((GEExcreta * %AIAFeed)/ %AIAExcreta) – 8.22 * (%NFeed – (%NExcreta * %AIAFeed / %AIAExcreta))

The N correction factor of 8.22 kcal/g of N retained was used according to formulas from Dias et al. (2023) and Xie et al. (2021).

Statistical analyses

Normality was confirmed using the Shapiro-Wilk test in the UNIVARIATE procedure in SAS 9.4 (P > 0.05; Cary, NC). Data were initially analyzed as a two-way ANOVA using the MIXED procedure and post hoc comparisons were conducted using Tukey-Kramer test if significant differences were detected (P < 0.05). Bodyweight on d 28 (beginning of the heat stress period) affected BW at 35 d of age and was therefore included as a covariate for that measure and analyzed using a two-way analysis of covariance. In trial 1 (floor pens), all data were analyzed on a pen basis, and each room was considered a block. For meat yield, meat quality, and the biomarkers, sex was also included as a factor. For trial 2 (bioassay cages), data were analyzed on a cage basis. If AA content significantly affected BWG and F:G (P < 0.05), the results were analyzed using linear broken line (LBL) and quadratic broken line (QBL) regressions using Proc NLIN with the independent variable being % of estimated AA requirements. The regression models and SAS procedures were performed according to Robbins et al. (2006):

The linear broken-line model used was:

$$y=L+U\times (R-x),\mathit{where}(R-x)\mathit{is}\mathit{defi}\mathit{ned}\mathit{as}\mathit{zero}\mathit{when}x>R.$$

The quadratic broken-line model used was:

$$y=L+U\times (R-x)*(R-x),\mathit{where}(R-x)\mathit{is}\mathit{defi}\mathit{ned}\mathit{as}\mathit{zero}\mathit{when}x>R.$$

Where y is the dependent variable (BWG or F:G), x is the dietary AA concentration, L is the response at the plateau, U is the slope and R is the AA concentration at the breakpoint. The following formula was used in Proc SQL to calculate R²:

$$R^2=\left(CSS--SSE\right)/CSS$$

Where CSS was the corrected total sum of squares, and SSE was the sum of squares of the error for the model.

Results and discussion

Growth performance

Growth performance results for Trial 1 are summarized in Table 3. No effects were seen on mortality during the trial. Initial BW at the start of the finisher period (d 21) was similar between treatments (P = 0.81). At d 28 and 35 there were no interactions observed between AA level and S:L ratio on BW (P = 0.49 and P = 0.91, respectively). At d 28 (start of heat stress), BW was increased with dietary AA content from 1.780 kg at 80% to 1.866 kg at 130% (P < 0.01). This trend carried over to BW on d 35 (end of heat stress and finisher period) increasing from 2.424 kg at 80% AAR to 2.586 kg in birds fed 130% (P < 0.01). There was a difference in BW at d 28 (P < 0.01) between the groups fed diets with S:L ratios of 4:1 and 10:1. Birds fed the 10:1 S:L ratio diets weighed an average of 1.853 kg, while those on the 4:1 S:L ratios diets averaged 1.817 kg. However, no effect of S:L on BW was observed on d 35 (P = 0.10).

Table 3.

Trial 1- Effects of dietary amino acid content and starch-to-lipid ratios on bodyweight, feed intake, feed conversion, and mortality of mixed sex Ross 308 broilers during the finisher period (d 21-35) with a 7-d period of cyclic heat stress (d 28-35)^1.

	Amino acid level (AA)						Starch-to-lipid ratio (SL)		Pooled SEM		P-Value
	80	90	100	110	120	130	4	10		AA	SL	AAxSL
Bodyweight (kg)
d 21	1.113	1.111	1.102	1.111	1.109	1.105	1.104	1.108	0.0020	0.78	0.59	0.73
d 28	1.780b	1.823ab	1.824ab	1.855a	1.861a	1.866a	1.817b	1.853a	0.0132	<0.01	<0.01	0.49
d 352	2.424b	2.509ab	2.522ab	2.578a	2.576a	2.586a	2.520	2.546	0.0252	<0.01	0.10	0.91
Bodyweight gain (kg)
d 21-28	0.667d	0.712c	0.722bc	0.744ab	0.752ab	0.761a	0.713b	0.745a	0.0142	<0.01	<0.01	0.03
d 28-35	0.644b	0.686ab	0.697ab	0.723a	0.716a	0.720a	0.703	0.692	0.0122	<0.01	0.37	0.99
d 21-35	1.312c	1.397b	1.419ab	1.467a	1.468a	1.481a	1.415	1.438	0.0262	<0.01	0.07	0.69
Feed intake (kg)
d 21-28	1.143a	1.133a	1.116ab	1.089ab	1.074ab	1.056b	1.086b	1.118a	0.0134	<0.01	<0.01	0.02
d 28-35	1.274a	1.260a	1.219ab	1.202ab	1.192ab	1.184b	1.221	1.222	0.0106	<0.01	0.90	0.74
d 21-35	2.417a	2.394a	2.336ab	2.291bc	2.266bc	2.241c	2.307b	2.341a	0.0227	<0.01	0.02	0.06
Mortality corrected feed-to-gain ratios (kg/kg)
d 21-28	1.715a	1.597b	1.539bc	1.463cd	1.425d	1.390d	1.534	1.509	0.0349	<0.01	0.22	0.09
d 28-35	1.972a	1.843ab	1.748bc	1.661c	1.670c	1.654c	1.741	1.772	0.0316	<0.01	0.10	0.86
d 21-35	1.838a	1.714b	1.637c	1.558d	1.540d	1.510d	1.640	1.638	0.0348	<0.01	0.97	0.31
Mortality (%)
d 21-28	0.8	1.1	1.8	1.2	1.4	0.9	1.1	1.3	0.126	0.37	0.35	0.56
d 28-35	0.5	1.0	0.9	1.4	1.3	2.1	1.2	1.3	0.154	0.09	0.73	0.93
d 21-35	1.4	2.2	2.7	2.5	2.6	3.0	2.2	2.6	0.204	0.27	0.40	0.62

¹n = 8 per amino acid level per starch-to-lipid ratio.

²Analyzed with covariate of d 28 BW (P < 0.01) bodyweight gain was not affected

^abcd Letters indicate differences between treatment.

There was an interaction between AAR and S:L ratios for BWG on d 21-28 (P = 0.03), where birds fed the 10:1 S:L ratio diets at AAR levels of 120 and 130% of estimated requirements had increased gain, with BWG of 0.781 and 0.789 kg, respectively. In contrast, birds fed diets with the same AAR levels of 120 and 130% but with a 4:1 S:L ratio gained 0.722 kg and 0.733 kg, respectively. However, no interaction between AA and S:L ratios were observed during d 28-35 (P = 0.99) or d 21-35 (P = 0.69) for BWG. Body weight gain improved with increasing AA, ranging from 0.644 kg to 0.723 kg during d 28-35 (heat stress) and from 1.312 kg to 1.481 kg during d 21 to 35 (P < 0.01 for the overall period). The S:L ratio did not affect BWG during d 28-35 (P = 0.37) or from d 21 to 35 (P = 0.07).

There was an interaction between AAR and S:L ratios for feed intake only for the d 21-28 period (P = 0.02) where feed intake was lower in birds fed a S:L of 4:1 and 120% AAR level compared to the treatment group with 10:1 S:L ratio (1.035 kg vs 1.113 kg respectively). The lower BWG observed in birds fed diets with an S:L of 4:1 was related to reduced feed intake during d 21-28. This was likely caused by the increased fat content of the low S:L ratio diets reducing pellet durability due to lubrication in the die during compaction, reducing temperature and starch gelatinization and increasing the proportion of fines (Bastiaansen et al., 2025; Dozier et al., 2010). Another reason for this difference could be related to the differences in diet compositions in terms of both nutrient and ingredients. The differences in diet composition in terms of fiber, protein, starch, and lipids likely affected the rate and efficiency of digestion and absorption, possibly causing changes in utilization (Bryan et al., 2019; Liu and Selle, 2015). No interactions between AAR and S:L ratios in the periods from d 28-35 and d 21-35 were observed (P = 0.74 and P = 0.06, respectively). Overall, feed intake exhibited a linear reduction with increasing dietary AA density during the finisher period, with higher feed intake in birds fed 80% AAR compared to those fed 130% AAR (P < 0.01). There was no effect of the S:L ratio on feed intake during the heat stress period (P = 0.90); however, feed intake of birds fed diets with S:L ratio of 10:1 on d 21-35 was higher (2.307 vs 2.341, P < 0.01).

There were no interactions between AAR and S:L ratios on F:G ratios in any of the evaluated periods. Furthermore, S:L ratios did not have an effect on F:G ratios during any of the assessed periods. However, there was a notable improvement in F:G ratios during all periods of the finisher phase with increased AAR. Birds fed a diet containing 130% AAR exhibited lower F:G ratios compared to those fed 80% AAR (P < 0.01). This trend was also observed in our previous trial, where we found lowest F:G ratios in birds fed the highest AA content diet (Duhra et al., 2025).

The data for BWG exhibited a relatively poor fit (R² ranging between 0.218-0.494) when analyzed using either LBL or QBL regression (Table 4). In particular, the fitted models for BWG during the heat stress period (d 28-35) had a R² of less than 0.216 and 0.218 for LBL and QBL respectively. Body weights and feed intakes were quite variable during the heat stress period. Mixed-sex Ross 308 broilers were utilized, and the observed variation may be attributed to the differing responses of males and females to elevated environmental temperatures. Male broilers are heavier and are more susceptible to the adverse effects of heat stress (Gogoi et al., 2021), while female broilers tend to have a lower growth rate but a greater capacity for energy storage in fat depots, making them less affected by high temperatures (Cahaner and Leenstra, 1992). During the thermoneutral period from d 21-28 the estimated breakpoint was 114% for LBL and 131% for QBL to achieve a BWG of 0.756 kg and 0.759 kg, respectively. During the heat stress period (d 28-35) the estimated value for LBL and QBL reduced to 107% and 116% likely due to increased energy requirements for maintenance in this period compared to d 21-28 (0.720 kg and 0.719 kg, respectively). However, a portion of this effect is likely due to increased energy requirements for maintenance in d 28-35 compared to d 21-28 due to a combination of increased BW and heat stress exposure reducing feed intake increasing the proportion of energy intake going towards maintenance and growth rates (Morillo et al., 2023; Teyssier et al., 2022). In the overall period (d 21-35) the recommendation would be 110% and 122% AAR based on LBL and QBL, respectively. The recommendation using QBL was generally higher despite the similar maximum BW between the two models. While the LBL can effectively describe the response of an individual animal, less accurately captures the dynamics of a population due to the abrupt transition at the plateau, which should ideally be a smooth curve (Wang et al., 2019). In this regard, the QBL would be better in describing the dynamics of the population compared to LBL since the ascending proportion of the model describe a curvilinear response up to the plateau while also providing a recommendation at the breakpoint (Pesti et al., 2009). This approach would ensure that most individuals receive adequate levels of essential AA, especially given the significant variation we observed in feed intake and BW.

Table 4.

Trial 1- Modeled effects of dietary amino acid density on growth performance broilers with a period of cyclic heat stress (d 28-35) during the finisher period (d 21-35).

Model	Period	Equations	R2	P-value	Breakpoint (AAR)
Body weight gain
	d21-28	Y = 0.756 – 0.0024*(114 - X)	0.458	<0.01	114
Linear broken line plateau1	d28-35	Y = 0.720– 0.0027*(107 - X)	0.216	<0.01	107
	d21-35	Y = 1.474 – 0.0049*(110 - X)	0.486	<0.01	110
	d21-28	Y = 0.759 – 0.000003*(131 – X)2	0.467	<0.01	131
Quadratic broken line1	d28-35	Y = 0.719 – 0.000006*(116 - X)2	0.218	<0.01	116
	d21-35	Y = 1.475 – 0.00009*(122 - X)2	0.494	<0.01	122
Feed to gain ratio
	d21-28	Y = 1.408 + 0.0096*(115 - X)	0.855	<0.01	115
Linear broken line plateau1	d28-35	Y = 1.659 + 0.011*(108 - X)	0.713	<0.01	108
	d21-35	Y = 1.525 + 0.001*(114 - X)	0.888	<0.01	114
	d21-28	Y = 1.388 – 0.00012*(136 - X)2	0.864	<0.01	136
Quadratic broken line1	d28-35	Y = 1.653 – 0.00019*(121 - X)2	0.711	<0.01	121
	d21-35	Y = 1.515 – 0.00014*(129 - X)2	0.895	<0.01	129

Abbreviation: AAR (Amino acid requirement).

¹the expression (R – x) is zero at values of x > R, where R is the breakpoint and x is the amino acid level.

On the other hand, F:G ratios exhibited less variation and were much more consistent, resulting in better fitting models at all age ranges. During the thermoneutral period from d 21-28, the estimated breakpoints were 115% for LBL and 136% for QBL to achieve F:G ratios of 1.408 and 1.388, respectively. During the heat stress period (d 28-35), the estimated AAR using LBL and QBL were 108% and 121%, with estimated F:G of 1.659 and 1.653, respectively. In the overall period (d 21-35), the recommended AA levels would be 114% and 129% based on LBL and QBL, respectively. Although the recommendations for BWG and F:G seems close, it may be more appropriate to adjust the AA levels according to the F:G ratio criteria, as the data showed a better fit. The recommendations align with the findings of a previous study (Maharjan et al., 2020), which suggested a 110 to 120% AA level for both optimal feed conversion ratio and average daily gain for heat stressed broilers from d 22-42 (∼29.6°C, 80.9% RH). However, as the current study estimated an AAR of 129% for the overall finisher period (d 21-35) this may not accurately represent the actual requirement of the animals, and further increased AA content may result in higher estimated requirements for this period.

Meat yield and quality

Meat yield and quality results are summarized in Table 5. Increasing AA content resulted in increased carcass and P. minor yields but reduced back and rack yields as a percent of live weight (P < 0.01 for all) which are indicative of increased muscle deposition with increased AA levels. Birds fed a S:L ratio of 10:1 had higher yields of carcass, whole thighs, wings, and back and rack yields as a percent of live weight than birds fed a diet with a S:L of 4:1 (P < 0.05 for all). Breast meat yield increased up to 23.1% of live weight in birds fed 120% AAR compared to 20.6% in birds fed 80% AAR (P < 0.01). However, breast yield was still lower than what would have been expected in birds of similar weight of 25.5% and 26.6% of liveweight males and females respectively under thermoneutral conditions (Aviagen, 2022). This is primarily due to reduced breast muscle deposition in both sexes (Teyssier et al., 2022) possibly related to increased maintenance costs associated with maintaining homeostasis taking priority metabolically during the heat stress. There were interactions for P. major and total breast yield between AAR and S:L ratios, and between S:L ratios and sex. Birds fed 80% AAR and a S:L of 4:1 had a higher P. major yield and breast yield of 17.4% and 21.0% of liveweight while birds fed 80% AAR and a S:L of 10:1 had a yield of 16.5% and 20.0% of liveweight for P. major and total breast yield respectively, however yield was similar between S:L ratios at 100 and 120% AAR (P < 0.05 for all). For the S:L and sex interaction males fed S:L of 4:1 had P. major and total breast yields of had increased yield of 18.5% and 22.1% compared to males fed diets with an S:L of 10:1 with yields of 17.9 and 21.5% for P. major and total breast yield respectively (P < 0.05 for all). The higher breast yield in the males fed a S:L ratio of 4:1 compared to 10:1 at 80% AAR could be due to the higher fat content of the diets allowing for improved AA digestion and absorption at low AA levels (Mateos et al., 1982), although the highest breast meat yield was observed in birds fed 120% AAR and an S:L ratio of 10:1 which may be indicative that these effects only occur at low AA levels. The increasing breast meat yield with AA content followed similar trends as our previous research where increasing inclusion of the high AA diet increased breast yield (Duhra et al., 2025) indicating that the increased AA content of the diets was being utilized for muscle deposition. Yield of P. minor were higher in females at 3.9% compared to 3.6% in males while they had lower yield of drumsticks (P < 0.05). As female broilers are expected to have higher breast meat yield the higher P. minor yield was expected (Aviagen., 2022). The difference in drumstick yields was expected based on Aviagen performance objectives (2022) and were similar for expected yields of 9.8% and 9.3% for males and females respectively whereas 9.7% and 9.5% was measured in trial birds. As such, the most affected muscle due to AA content was the breast, due to AA content, which was expected based on previous research (Duhra et al., 2025; Teyssier et al., 2022).

Table 5.

Trial 1. Effects of dietary amino acid content and starch-to-lipid ratios on carcass characteristics of 36-d-old male and female Ross 308 broilers following a 7-d period of cyclic heat stress (d 28-35)^1.

	Amino acid level (AA)			Starch-to-lipid ratio (SL)		Sex		Pooled SEM	P-value
	80	100	120	4	10	Male	Female		AA	SL	Sex	AAx SL	AAx Sex	SLx Sex	AAx SLx Sex
Percent of live weight
Carcass	71.7b	73.1a	73.5a	72.1b	73.4a	72.7	72.8	0.44	<0.01	<0.01	0.86	0.97	0.75	0.95	0.18
Pectoralis major	17.0c	18.3b	19.1a	18.2	18.1	18.2	18.1	0.42	<0.01	0.63	0.43	0.04	0.94	0.03	0.09
Pectoralis minor	3.6b	3.9a	3.9a	3.8	3.8	3.6b	3.9a	0.07	<0.01	0.97	<0.01	0.33	0.27	0.77	0.29
Total breast	20.6c	22.2b	23.1a	22.0	21.9	21.9	22.0	0.34	<0.01	0.52	0.47	0.02	0.94	0.03	0.05
Whole thighs	13.2	13.5	13.3	13.1b	13.5a	13.3	13.3	0.10	0.10	<0.01	0.48	0.08	0.40	0.80	0.13
Drums	9.6	9.5	9.6	9.5	9.6	9.7a	9.5b	0.05	0.84	0.49	0.03	0.15	0.50	0.66	0.49
Wings	7.5	7.5	7.5	7.4b	7.6a	7.5	7.5	0.04	0.71	0.03	0.26	0.63	0.99	0.45	0.77
Breast skin	1.9a	1.8a	1.7b	1.7b	1.9a	1.8	1.8	0.05	<0.01	<0.01	0.41	0.51	0.73	0.32	0.03
Back and rack	18.7a	18.4ab	18.2b	18.1b	18.8a	18.4	18.5	0.19	<0.01	<0.01	0.69	0.29	0.44	0.32	0.18
Meat quality measures
Initial pH	7.26	7.31	7.31	7.31	7.28	7.30	7.29	0.016	0.34	0.31	0.60	0.66	0.02	0.42	0.84
Final pH	6.05	6.09	6.12	6.06	6.11	6.08	6.09	0.023	0.37	0.13	0.88	0.02	0.42	0.18	0.41
ΔpH	1.22	1.22	1.20	1.24	1.17	1.22	1.20	0.024	0.91	0.11	0.77	0.06	0.85	0.15	0.82
L*	50.43	49.65	49.23	50.32	49.30	50.54a	49.03b	0.341	0.16	0.07	<0.01	0.28	0.98	0.81	0.64
a*	4.06	3.76	4.25	4.17	3.89	3.85	4.19	0.164	0.19	0.19	0.12	<0.01	0.69	0.51	0.42
b*	6.90a	5.67b	3.63c	4.58b	6.20a	5.42	4.37	0.553	<0.01	<0.01	0.95	<0.01	0.29	0.92	0.71
Change in weight (%)
Drip loss	1.22a	0.79b	0.79b	0.96	0.91	0.83	1.04	0.093	0.02	0.79	0.14	0.60	0.09	0.58	0.38
Thaw loss	4.34	3.86	3.98	4.07	4.04	3.86	4.25	0.207	0.34	0.98	0.17	<0.01	0.47	0.03	0.87
Cook loss	21.38	21.04	20.01	20.81	20.81	21.32	20.30	0.347	0.16	0.99	0.10	0.07	0.39	0.51	0.97

¹n = 8 per AA level per starch-to-lipid ratio per sex

^abc Letters indicate differences between treatments.

Breast skin yield was highest in females fed 80% AAR and a S:L ratio of 10:1 and lowest in males fed 120% AAR and S:L ratio of 4:1. As the birds fed the 10:1 S:L ratio diets had higher breast skin yield regardless of sex, it appears that the higher proportion of energy coming from starch resulted in increased lipid deposition, although the exact reason for this is unclear, it may be related to a surplus of energy in birds fed high S:L ratio diets with a low AA density. Previous research has shown that diets with a lower nutrient density but a higher S:L ratio had higher fat deposition on the carcass which may have led to higher skin yield observed in this study in females fed 80% AAR and a S:L ratio of 10:1 (Khoddami et al., 2018).

There was an interaction between sex and AA content for P. major pH where males fed a diet with a 100% AA content had a higher initial pH of 7.31 relative to females fed diets with the same AA content (7.26, P < 0.05). The small but significant difference in initial pH between males fed diets with S:L ratios of 4:1 and 10:1 did not carry over to final pH or change in pH, thus it was likely not biologically significant. An interaction between AAR and S:L ratio was observed where birds fed diets with an AAR of 80% and a S:L ratio 10:1 had a higher final P. major pH than those fed a diet with an S:L of 4:1 at the same AAR (6.15 vs 5.96). However, the initial and change in pH were not different between the treatments and thus this may be just a random effect (P > 0.05). For L* measurements (lightness), males had a higher reading of 50.54 than females which averaged 49.03 (P < 0.01), indicating the muscle was lighter in color for males. The change in light color was likely due to more rapid postmortem anaerobic glycolysis, producing lactic acid, in the chicken breast which could be indicative of increased potential for pale-soft-exudative meat in birds (Wang et., 2017). However, the difference was small and within the expected range of 46 to 53, thus not likely to be biologically significant (Barbut et al., 2005). There was an interaction between AA content and S:L ratios on a* (redness) where bird fed diets with 120% AAR and a S:L of 10:1, or diets with AA content of 80 and 100% with a S:L ratio of 4:1 had higher a* values than birds fed a diet with AAR of 100% and a S:L ratio of 10:1, however the cause of this is unclear. For b* (yellowness), breast meat decreased with increasing AA content, decreasing from 6.90 at 80% AA to 3.63 at 120% AAR (P < 0.01). Breast muscle yellowness was also lower in birds fed diets with 120% AAR at both S:L ratios, and birds fed 100% AAR at an S:L of 4:1 compared to the other diets. The diets with 80 and 100% AAR with a S:L ratio of 10:1 likely had higher b* values due to the high inclusion of corn which has been observed in previous research to affect pigmentation of meat due to high carotenoid content (Smith et al., 2002). While this explains the pigmentation observed in birds fed diets containing 70% or more corn, the reason for the birds fed 80% AAR and a S:L ratio of 4:1 having elevated b* is uncertain. Drip loss was lower in birds fed 100 and 120% AAR at 0.79% compared to birds fed 80% AAR which had a drip loss of 1.22% (P = 0.02). This could be indicative of increased protein denaturation, reducing water holding capacity (Traore et al., 2012) in the P. major in birds fed diets with 80% AAR. There were interactions between AA content and S:L ratios, and sex and S:L ratios for thaw loss. Thaw loss was higher in the P. major of broilers fed 80% AAR and a S:L ratio of 4:1 (5.00%) compared to birds fed 100% AAR and a S:L ratio of 10:1 (3.36%) while other treatments were similar. As the birds fed the other treatments had similar thaw loss to the two treatments, the exact cause of this is unknown and may be due to random effect. For the interaction between sex and S:L ratios, P. majors of male birds fed diets with a S:L ratio of 4:1 had lower thaw loss than females fed the same S:L (3.54 vs 4.58%), but both were similar to thaw loss of birds fed a S:L ratio of 10:1 (4.17 and 3.92% for males and females respectively), potentially indicating another random effect.

Oxidative stress biomarkers and bursa relative weights

Results for TBARS, catalase, glutathione reductase and bursa weights (as a percentage of live weight; collected on d 31) are summarized in Table 6. The TBARS were used as a measure of oxidative damage. Catalase and glutathione reductase were used as indicators of antioxidant capacity. Bursa weights were used as a measure of immune function, as heat stress has been shown to reduce the proportional weight of the bursa relative to BW (Chen et al., 2024). Overall, no interactions were observed between the dietary effects of AA levels, S:L ratios, or sex, nor their individual effects on the biomarker measurements. Bursa relative weights were not affected by sex, S:L ratios, or AA levels during heat stress in this experiment indicating that the diet compositions did not affect immune response under the simulated environmental conditions. Due to the nature of cyclic heat stress, catalase and glutathione peroxidase would be expected to increase after the heat stress period as a response against reactive oxygens species such as superoxide which can vary by durations and severity of heat stress (Akbarian et al., 2016). As this model had 12 h periods at a comfortable temperature the birds may have had enough time to recover between heat stress periods, and, as the samples were collected during the first few hours of the heat stress period, it may not have been possible to detect effects of AA levels or S:L ratios in male or female broilers on the biomarkers evaluated. In addition to the previously discussed potential impact of the sample collection period on the lack of a significant response, it is important to mention that assays for TBARS also have low specificity and can react with breakdown products of protein and carbohydrates which may have influenced the measure (Ghani et al., 2017). There were indications, such as the reduction in drip loss with increasing AA levels, suggesting that testing specifically for protein oxidation could provide meaningful insights into the potential correlation with meat quality and moisture loss under the simulated environmental conditions (Traore et al., 2012). As heat stress increases oxidative damage, alterations in amino acid utilization may explain the lower breast meat deposition, as AA are likely redirected for maintenance needs (Temim et al., 2000). Therefore, measuring markers such as creatine kinase could also be beneficial in future research.

Table 6.

Trial 1- Effects of dietary amino acid level, starch-to-lipid ratios, and sex on bursa and breast and liver biomarkers of 31-d-old heat stressed Ross 308 broilers (31.0°C, 50% RH)^1,2,3.

	Amino acid level (AA)						Starch-to-lipid ratio (SL)		Sex		Pooled SEM	P-value
	80	90	100	110	120	130	4	10	Male	Female		AA	SL	Sex	AAx SL	AAx Sex	SLx Sex	AAxSLx Sex
BW (kg)	2.13	2.18	2.17	2.25	2.21	2.23	2.20	2.20	2.37a	2.03b	0.038	0.10	0.80	0.01	0.97	0.44	0.55	0.35
Bursa (%BW)	0.15	0.16	0.14	0.16	0.15	0.16	0.16	0.15	0.16	0.15	0.003	0.25	0.26	0.29	0.66	0.63	0.62	0.26
Pectoralis major biomarkers
TBARS (µM)	0.052	0.046	0.049	0.031	0.048	0.042	0.042	0.048	0.043	0.047	0.003	0.68	0.44	0.59	0.43	0.94	0.13	0.73
GPX (nmol/min/ml)	52.7	49.8	51.1	51.6	44.7	47.8	47.9	51.4	48.9	48.9	1.19	0.76	0.29	0.69	0.91	0.55	0.64	0.80
CAT (nmol/min/ml)	27.5	26.9	24.3	24.7	26.9	21.8	24.9	25.7	25.7	24.9	0.93	0.73	0.76	0.75	0.53	0.59	0.37	0.74
Liver biomarkers
TBARS (µM)	0.777	0.567	1.593	0.766	0.619	0.773	0.680	1.019	0.695	1.002	0.1774	0.63	0.36	0.41	0.44	0.39	0.25	0.47
GPX (nmol/min/ml)	120.2	110.4	105.2	107.9	102.0	103.0	109.7	106.5	108.4	107.9	2.46	0.26	0.50	0.92	0.17	0.49	0.76	0.42
CAT (nmol/min/ml)	136.8	126.7	122.9	135.5	138.8	123.9	133.7	127.8	130.6	130.9	2.88	0.45	0.33	0.95	0.91	0.07	0.51	0.59

Acronyms: TBARS (thiobarbituric acid reactive substances), GPX (glutathione peroxidase), CAT (catalase), AA (amino acid), SL (starch-to-lipid ratio).

¹n = 8 per amino acid level per starch-to-lipid ratio per sex.

²Interassay coefficient of variation <10% for all.

³Intraassay coefficient of variation <10% for all

^ab Letters indicate differences between treatments.

Nutrient retention

Nutrient retention results obtained from the excreta collected on d 32 to 34 and growth performance results of the Trial 2 animals are summarized in Table 7. Bodyweight on d 28 was affected by the S:L ratios with birds fed diets with a S:L of 4:1 having lower BW of 1.881 kg compared to 1.927 kg in birds fed diets with an S:L of 10:1 (P = 0.02). This was likely related to the lower feed intake of birds of 1.147 kg in birds fed S:L 4:1 compared to 1.216 kg in birds fed diets with a S:L of 10:1 (P < 0.01). Feed intake was reduced from d 21-28 by increasing AA content where birds fed 120% AAR had lower feed intake than birds fed 80% AAR (1.148 vs 1.253 kg respectively; P = 0.01). Dietary amino acid content did not affect feed intake (d 28-35 and d21-35), BW, BWG, or mortality for the overall finisher period in the Trial 2 birds (P > 0.05). Feed to gain ratios were reduced by increasing AA content of the diets during all periods during the finisher phase similar to what was observed in the Trial 1 animals (P < 0.05). Both DM and N retention was highest in birds fed 90 and 100% AAR content and lowest in birds fed 110% AA (P < 0.05) which could be indicative of a metabolic shift related to the increasing AA content which reduced nutrient retention, possibly due to AA being used for energy increasing N excretion as a waste product (Maeda et al., 2017). Measured AMEn was unaffected by dietary AA content (P > 0.05). Dry matter retention was higher for birds fed a S:L ratio of 10:1, likely due to the 4:1 S:L ratio diets containing wheat middlings while the 10:1 S:L ratio diets did not. Wheat middlings have lower DM digestibility compared to ingredients like corn and soybean meal (Adedokun et al., 2015; Oviedo-Rondón et al., 2024). Proximate analyses of diets with S:L ratio of 4:1 indicated higher crude fiber content of approximately 3.7% on average compared to 2.8% in 10:1 S:L ratio diets. Diets with a S:L ratio of 4:1 had higher acid detergent fiber of 5.1% compared to 3.9% and neutral detergent fiber of 13.2% compared to 9.6%. In particular, the higher neutral detergent fiber content may have increased digesta viscosity reducing nutrient digestibility and absorption (Adedokun et al., 2015) although it was a relatively small increase as a percent of the diet.

Table 7.

Trial 2- Effects of dietary amino acid content and starch-to-lipid ratios on growth performance and nutrient retention of mixed sex Ross 308 broilers in bioassay cages during the finisher period (d 21-35) with a 7-d period of cyclic heat stress (d 28-35)^1.

	Amino acid level (AA)2						Starch-to-lipid ratio (SL)		Pooled SEM	P-Value
	80	90	100	110	120	130	4:1	10:1		AA	SL	SLxAA
Bodyweight (kg)
d 21	1.118	1.092	1.097	1.112	1.114	1.113	1.101	1.131	0.0111	0.45	0.16	0.36
d 28	1.860	1.883	1.899	1.919	1.933	1.941	1.881b	1.927a	0.0191	0.13	0.02	0.31
d 353	2.497	2.494	2.574	2.619	2.597	2.589	2.552	2.587	0.0210	0.20	0.39	0.65
Bodyweight gain (kg)
d 21-28	0.743	0.791	0.802	0.807	0.818	0.828	0.779	0.816	0.0110	0.39	0.13	0.88
d 28-35	0.637	0.611	0.675	0.700	0.664	0.648	0.671	0.640	0.0110	0.37	0.20	0.84
d 21-35	1.379	1.402	1.477	1.507	1.483	1.476	1.451	1.456	0.0161	0.33	0.89	0.86
Feed intake (kg)
d 21-28	1.253a	1.223ab	1.178ab	1.149ab	1.139b	1.148ab	1.147b	1.216a	0.0183	0.01	<0.01	0.30
d 28-35	1.214	1.150	1.196	1.180	1.146	1.130	1.193	1.145	0.0140	0.53	0.11	0.63
d 21-35	2.467	2.374	2.373	2.329	2.285	2.278	2.341	2.361	0.0227	0.15	0.63	0.70
Mortality corrected feed-to-gain ratios
d 21-28	1.689a	1.549ab	1.487b	1.430b	1.403b	1.399b	1.484	1.501	0.0320	<0.01	0.59	0.68
d 28-35	2.000a	1.947a	1.814ab	1.763b	1.766b	1.792ab	1.801	1.892	0.0336	0.02	0.07	0.65
d 21-35	1.825a	1.729ab	1.630bc	1.580c	1.551c	1.558c	1.626	1.661	0.0308	<0.01	0.13	0.67
Mortality (% of birds)
d 21-28	0.00	0.00	3.13	0.00	0.00	0.00	1.04	0.00	0.521	0.43	0.32	0.43
d 28-35	3.13	3.13	3.13	3.13	3.13	3.13	2.08	4.17	0.942	0.99	0.42	0.39
d 21-35	3.13	3.13	6.25	3.13	3.13	3.13	3.12	4.17	0.929	0.98	0.71	0.54
Nutrient retention (%; d 32-34)
Dry matter	74.8ab	76.2a	76.2a	73.3b	75.1ab	74.9ab	71.7b	78.4a	0.44	0.03	<0.01	0.54
N retention	66.7ab	68.5a	67.2ab	60.3c	64.7abc	62.1bc	65.2	64.7	1.30	<0.01	0.64	0.91
AMEn (kcal/kg)	3,025	3,084	3,078	3,003	3,071	3,049	3,029b	3,074a	13.1	0.11	0.02	0.33
Total energy intake (kcal/bird)
d 28-35	3,672	3,548	3,682	3,542	3,517	3,447	3,613	3,523	37.5	0.67	0.34	0.57
Energy-intake-to-BW-gain ratio (kcal/kg BW gain)
d 28-35	6,057a	6,001ab	5,580bc	5,294c	5,428bc	5,497bc	5,460b	5,815a	129.5	0.03	0.02	0.53

Abbreviation: AMEn (nitrogen corrected apparent metabolizable energy).

¹Mean of replicates n = 8 per treatment.

²Percent of estimated requirement.

³Affected by bodyweight at 28 days of age so day 28 bodyweight was included as a covariate

^abc Letters indicate differences between treatments.

Energy intake was not affected by AA or S:L ratios (d 28-35; P = 0.67). Energy-intake-to-BWG (E:BWG) ratios was lowest in birds fed 110% AAR and highest in birds fed 80% during the same period. As birds fed 110% AAR had the lowest E:BWG ratio, this could be indicative of AA requirements being met. CP retention was also lowest in birds fed 110% AAR, however the reduction in N retention may have been related to a reduction in efficiency of digestion and absorption of N due to the increased supply of AA. In particular, the increasing supplementation of synthetic AA as dietary AA levels increased may have resulted in reduced efficiency of absorption of non-bound AA (Macelline et al., 2025). The overall reduction in E:BWG ratios associated with increasing AA content is likely related to increased muscle deposition. Protein deposition is more energetically efficient than fat deposition per gram of deposited material (8.6 vs 11.1 kcal ME/g protein or fat deposited, respectively; Barzegar et al., 2020). Lean muscle mass is also approximately 75% water by weight compared to approximately 10% in body fat (Lorenzo et al., 2019) and thus muscle requires significantly less energy to deposit by weight compared to fat, leading to the improved E:BWG and F:G ratios observed during both trials with increasing AA content. The birds fed the 10:1 S:L ratio diets had higher energy intake to BWG ratios during d 28-35 (P = 0.02), which could explain why final BW and BWG (28-35) was similar between the S:L ratios. This could be related to improved nutrient utilization during heat stress due to the higher fat content of the diets improving nutrient retention by reducing passage rates (Mateos et al., 1982), however AMEn was lower in birds fed a S:L of 4:1 which appears to counter the point. Another reason could be differences in digestion rates between the diets due to differences in the fiber, starch, lipid, and oil contents of the diets, resulting in variations in nutrient utilization (Bryan et al., 2019; Mateos et al., 1982; Tejeda and Kim, 2021). Previous research on pair-fed birds has shown that heat stressed birds consuming the same amount of feed as those in thermoneutral conditions presented lower body weights and poorer feed efficiency indicating alterations in metabolism (Teyssier et al., 2022). A future direction for research could entail a pair feeding trial using broilers reared under thermoneutral conditions are pair fed the intake of heat stressed treatment, and assess the effect n terms of growth, nutrient retention, and carcass composition. Further research could entail improving nutrient availability thus improving nutrient utilization and reducing excretion which could improve efficiency and sustainability.

The overall recommendations to meet requirements for F:G ratios and BWG would be met by formulating for AAR of 114 and 129% using LBL and QBL respectively. However, feed is the largest cost of production (Son et al., 2024), as such consideration must be taken regarding how economical it is to alter formulations and increase costs. As increasing AA content of diets is expensive when considering meeting those requirements using ingredients like soybean meal, and synthetic AA, the increased costs may not offset the improvements to performance in term of BWG, F:G, or meat yield depending on targets. In the case of breast yield, yield increased by 1.6% of liveweight from 80 to 100% AAR but only 0.9% from 100 to 120%. As such, the benefits in terms of production may not be offset by increased costs, potentially resulting in lower profit for stakeholders. This may vary based on prices of feed and the final product (BW, carcass, meat yield). Other challenges may include trying to meet certain AA targets. l-leucine at the time of this research was not approved for use in feed in Canada. As such, other ingredients need to be used to meet those requirements which may result in higher CP content causing a similar increase in variation when formulating for high increased AA levels. Another consideration is how much of a tangible benefit is observed when feeding these recommended diets compared to using primary breeder recommendations. This last point is a potential next step in research which may help guide recommendations for industry.

In conclusion, dietary S:L ratios ranging from 4:1 and 10:1 did not significantly impact broiler BWG and F:G ratios during either heat stress (d 28-35) or the overall finisher period (d 21-35). Increasing dietary digestible AA content to approximately 110-122% of the requirement maximizes BWG, while increasing dietary AA content to 114-129% optimized F:G ratios during the finisher phase (d 21-35), which included a period of cyclic heat stress (12 h 31 °C; 12 h 21 °C, RH 50-60%) from d 28 to 35. A higher digestible AA content would not only be more suitable to meet the requirement of a population of broilers but also improve breast meat yield if the higher recommendation by quadratic broken line is utilized.

CRediT authorship contribution statement

Dilshaan Duhra: Writing – review & editing, Writing – original draft, Methodology, Formal analysis. Denise Beaulieu: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization. Tory Shynkaruk: Writing – review & editing, Methodology. Juliano C. de Paula Dorigam: Writing – review & editing, Visualization, Methodology, Formal analysis, Conceptualization. Rose Whelan: Writing – review & editing, Visualization, Methodology. Karen Schwean-Lardner: Writing – review & editing, Visualization, Supervision, Methodology, Investigation, Funding acquisition, Conceptualization.

Disclosures

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Karen Schwean-Lardner reports financial support was provided by Evonik Operations GmbH. Karen Schwean-Lardner reports financial support was provided by Natural Sciences and Engineering Research Council of Canada. Juliano C. de Paula Dorigam reports a relationship with Evonik Operations GmbH that includes: employment. Rose Whelan reports a relationship with Evonik Operations GmbH that includes: employment. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.