Interactive effects of dietary apparent metabolizable energy, crude protein, and lysine levels on growth and slaughter performance of starter Pekin ducks

Abstract

A 4 × 3 × 3 factorial experiment was conducted to evaluate the interactive effects of dietary AME (2,750, 2,900, 3,050, and 3,200 kcal/kg), CP (17, 18.5, and 20%), and Lys (0.8, 1.05, and 1.3%) levels on growth and slaughter performance of starter Pekin ducks (1–14 d). A total of 2,304 one-day-old White Pekin ducks were randomly assigned to 36 treatments (4 replicates of 16 birds). At 14 d, ADG, ADFI, FCR, abdominal fat, breast meat, and leg meat yields were measured. As a result, increasing dietary Lys from 0.8 to 1.05% increased breast and leg muscle yields (P < 0.0001). CP × Lys interactions were observed for ADG and ADFI (P < 0.0001). Increasing CP improved ADG at 1.05 and 1.3% Lys, but reduced ADG and ADFI under Lys deficiency (0.8%). An AME × Lys interaction was found for ADFI (P = 0.0447). Increasing AME had no effect at 0.8% Lys, but reduced ADFI at 1.05 and 1.3%Lys. Abdominal fat percentage was affected by AME × CP (P = 0.0041) and AME × Lys (P = 0.0017) interactions. At 3,050 and 3,200 kcal/kg AME, increasing CP or Lys reduced abdominal fat, whereas no effects were observed at 2,750 kcal/kg AME. A three-way interaction among AME, CP, and Lys affected FCR (P = 0.0077), where increasing CP increased FCR at 2,750 and 2,900 kcal/kg AME with 0.8% Lys, but decreased FCR at 2,750 kcal/kg AME with 1.3% Lys. Quadratic models indicated optimal levels of 2,995 kcal/kg AME, 18.65% CP, and 1.20% Lys for maximum ADG, and 3,199 kcal/kg AME, 19.58% CP, and 1.15% Lys for minimum FCR. In conclusion, dietary AME, CP, and Lys exhibited interactive effects on growth and slaughter performance of starter Pekin ducks.

Introduction

The continuous rise in feed ingredient prices has made feed cost the most significant expenditure in poultry production, accounting for approximately 70% of total operating costs (Tata et al., 2024; Mun et al., 2025). Consequently, precision nutrition has gained increasing attention within the industry. Precision nutrition aims to optimize animal growth performance, enhance product quality, and minimize feed wastage and environmental pollution (Zhang et al., 2020).

In modern poultry precision nutrition systems, apparent metabolizable energy (AME), crude protein (CP), and lysine (Lys) are three core nutrients determining growth performance and carcass quality. Energy serves as the foundation for all physiological activities in animals (Wang et al., 2025). Dietary CP has been regarded as the foundation for muscle growth and protein deposition (An and Kong, 2023). Lys, as the second limiting amino acid for poultry, has been widely demonstrated to play a central regulatory role in skeletal muscle protein synthesis (Jin et al., 2020; Suen et al., 2025). It has been shown that increasing dietary energy, protein, or Lys levels can improve production performance of poultry (Zhou et al., 2017; Cho et al., 2020; Hong et al., 2022). However, these nutrients do not function independently within the body but instead exhibit interactive effects. Wen et al. (2017) found that dietary energy levels affect Lys requirements, as dietary AME increased from 2,875 to 3,050 kcal/kg, the Lys requirement of ducklings also rose from 0.94 to 0.98%. Ashour et al. (2023) demonstrated that the interaction between dietary energy and protein affects the laying performance of quails from 8 to 20 weeks of age. Other studies indicated that under different protein levels, increasing dietary AME has different effects on laying hens' egg production rate (Reid and Maiorino, 1980). At medium-high protein levels (16 and 18%), AME enhances egg production performance, whereas at low protein levels (14%), it reduces the egg production rate. Yao Yu et al. (2024) further reported an interaction between dietary CP and AME for fecal nitrogen content in broilers, with the highest content observed in diets with 23% CP and 3,000 kcal/kg AME, and the lowest at 20% CP and 2,900 kcal/kg AME. Furthermore, the digestion and absorption of proteins, transmembrane transport of amino acids, and synthesis of proteins require energy (Marchingo and Cantrell, 2022; Dehnavi et al., 2025), whereas insufficient energy forces the body to utilize proteins and amino acids as energy (Smith et al., 2011). Compared to carbohydrates or fats, utilizing amino acids for energy is less efficient and more costly. This inefficiency stems from the fact that generating energy from amino acids consumes raw materials required for protein synthesis, increases nitrogen excretion, and requires additional energy for uric acid synthesis (Macelline et al., 2025). Therefore, balancing the ratios of energy, protein, and Lys in diets is essential for achieving optimal poultry performance and reducing production costs.

Previous studies have predominantly focused on the effects of single or dual nutrient levels on Pekin ducks, while systematic research on the interactions among AME, CP, and Lys remains insufficient. This limitation has, to some extent, constrained our ability to formulate precise dietary formulations. Therefore, this study aims to investigate the interactions among AME, CP, and Lys, systematically evaluate their effects on the growth and slaughter performance of Pekin ducks, and determine the requirements of AME, CP, and Lys for Pekin ducks during the brooding phase (1-14 d).

Materials and methods

All experimental procedures were approved by the Animal Welfare Committee of the Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, and performed in accordance with the guidelines for animal experimentation established by the National Institute of Animal Health.

Birds and housing

A 4 × 3 × 3 factorial experiment, using 4 dietary AME levels (2,750, 2,900, 3,050, and 3,200 kcal/kg), 3 dietary protein levels (17, 18.5, and 20%) and 3 dietary Lys levels (0.80, 1.05, and 1.30%), was conducted to investigate the effects of different levels of AME, CP and Lys on the growth and slaughter performance of starter Pekin ducks. The 36 experimental diets were formulated based on 12 basal diets, which were subsequently supplemented with L-Lys HCl to achieve the target Lys levels. According to the Chinese national standard "Nutrient requirements of meat-type duck" (SAC/TC 274, 2024; GB/T 45103-2024), the starter phase of Pekin ducks is defined as 1–14 days of age; thus, the experiment was conducted from hatch to 14 days. A total of 2,304 one-day-old White Pekin ducks (S1 improved line derived from Pekin duck) were obtained from Inner Mongolia Saifeiya Agricultural Technology Development Co., Ltd. (Inner Mongolia, China), with a similar average body weight (55 ± 1 g). Birds were not sexed, and sex was not included as an experimental factor. The birds were randomly assigned to 36 dietary treatments, each containing four replicate pens with 16 birds per pen. These birds were kept in plasticwire-floor pens in an environmentally controlled experimental building. The temperature was kept at 33°C (at floor level) from 1 to 3 d of age, and then it was reduced gradually to 28°C until 14 d of age. All birds had free access to water and feed and lighting was continuous. Water was provided by drip-nipple water supply lines and feed was fed in pellet form.

Experimental diets

A total of 12 basal diets were formulated with different AME (2,750, 2,900, 3,050, and 3,200 kcal/kg) and CP (17, 18.5, and 20%) levels, using corn, soybean meal, peanut meal, distillers dried grains with solubles (DDGS), and corn gluten meal as the primary ingredients (Table 1). The CP contents of these raw materials were measured, while other nutritional components were referenced to the "Nutrient requirements of meat-type duck" (SAC/TC 274, 2024; GB/T 45103-2024). The Lys levels in all basal diets were formulated to 0.8% (Table 1). On this basis, crystalline L-Lys HCl (amino acid content 78%) was added at 0, 3.2, and 6.4 g/kg to prepare experimental diets with Lys levels of 0.8, 1.05, and 1.3%, respectively.

Table 1.

Ingredient composition and nutrient contents of the basal diets (% as-fed basis).

Ingredient	Diet
	1	2	3	4	5	6	7	8	9	10	11	12
Corn (7.34%1)	60.08	57.04	54.00	65.38	62.36	59.30	67.20	63.60	59.94	67.20	63.70	60.30
Soybean meal (43.45%1)	16.00	16.00	16.00	15.00	15.00	15.00	15.00	15.00	15.00	9.60	9.00	8.40
Peanut meal (51.84%1)	7.36	10.91	14.44	7.41	10.93	14.50	8.05	11.65	15.28	7.30	11.36	15.40
DDGS (23.92%1)	6.00	6.00	6.00	6.00	6.00	6.00	4.00	4.00	4.00	4.00	4.00	4.00
Soybean oil							1.39	1.58	1.78	3.27	3.45	3.61
Corn gluten meal (63.64%1)										4.00	4.00	4.00
Dicalcium phosphate	1.08	1.00	0.95	1.05	1.00	0.95	1.10	1.05	1.00	1.12	1.10	1.00
Limestone	1.38	1.40	1.40	1.40	1.40	1.40	1.38	1.37	1.39	1.42	1.41	1.42
Salt	0.30	0.30	0.30	0.30	0.30	0.30	0.30	0.30	0.30	0.30	0.30	0.30
L-Lys HCl, 78%	0.10	0.05		0.12	0.07	0.01	0.12	0.07	0.02	0.27	0.23	0.19
DL-Met	0.19	0.18	0.17	0.19	0.18	0.17	0.19	0.18	0.18	0.17	0.17	0.16
L-Thr	0.13	0.10	0.07	0.13	0.10	0.07	0.14	0.11	0.08	0.17	0.14	0.12
L-Trp	0.03	0.01		0.03	0.02		0.03	0.02		0.05	0.04	0.03
L-ILe	0.09	0.05	0.02	0.10	0.06	0.02	0.10	0.07	0.03	0.13	0.10	0.07
Premix2	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Rice hulls	6.26	5.96	5.65	1.89	1.58	1.28
Total	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00
Calculated composition3
AME (kcal/kg)	2,750	2,750	2,750	2,900	2,900	2,900	3,050	3,050	3,050	3,200	3,200	3,200
CP (%)	17.00	18.50	20.00	17.00	18.50	20.00	17.00	18.50	20.00	17.00	18.50	20.00
Lys (%)	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80
Met (%)	0.45	0.45	0.45	0.45	0.45	0.45	0.45	0.45	0.45	0.45	0.45	0.45
Thr (%)	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72
Trp (%)	0.20	0.20	0.20	0.20	0.20	0.20	0.20	0.20	0.20	0.20	0.20	0.20
ILe (%)	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72	0.72
Ca (%)	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80	0.80
NPP (%)	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35
Analyzed composition4
CP (%)	17.89	19.10	20.32	17.59	18.77	20.21	17.28	18.27	20.23	17.38	18.73	20.42
Lys (%)	0.94	0.91	0.88	0.89	0.90	0.80	0.89	0.86	0.77	0.90	0.84	0.84

¹Crude protein content, analyzed value.

²Supplied the following per kilogram of total diet: Cu, 10 mg; Fe, 60 mg; Zn, 60 mg; Mn, 80 mg; Se, 0.3 mg; I, 0.2 mg; Cr, 0.15 mg; choline chloride, 1,000 mg; vitamin A, 10,000 IU; vitamin D₃, 3,000 IU; vitamin E, 20 IU; vitamin K₃, 2 mg; thiamin, 2 mg; riboflavin, 8 mg; pyridoxine hydrochloride, 4 mg; cyanocobalamin, 0.02 mg; calcium-d-pantothenate, 20 mg; nicotinic acid, 50 mg; folic acid, 1 mg; biotin, 0.2 mg.

³These values were calculated.

⁴These values were analyzed.

Abbreviations: AME, apparent metabolizable energy; CP, crude protein; DDGS, distillers dried grains with solubles; NPP, non-phytate phosphorus; Lys, lysine; Met, methionine; Thr, threonine; Trp, tryptophan; ILe, isoleucine.

Growth performance

The initial body weight of each pen of ducks was measured. During the experiment, the mortality of ducks was recorded daily and feed consumption was adjusted accordingly. After a 12-hour fast on day 14 of age, the ducks from each replicate were weighed. Growth performance was assessed by the average daily gain (ADG), average daily feed intake (ADFI), and feed conversion ratio (FCR) of Pekin ducks.

Slaughter performance

At the end of the experiment, the ducks were fasted for 12 hours. Two ducks close to average weight from each pen were selected and stunned electrically and then killed by neck cut. Subsequently, the abdominal fat, breast meat (including pectoralis major and pectoralis minor) and leg meat (including thigh and drum stick) were all removed manually from carcasses and weighed, and the percentages relative to live body weight at processing were also calculated. Breast and leg meat were all skinless and boneless.

Statistical analyses

Statistical analysis was performed using RStudio software (Version 4.5.2, Posit PBC, Boston, MA, USA). Pen served as the experimental unit. Data for growth and slaughter performance were subjected to a three-way ANOVA. Dietary AME, CP, and Lys levels and their interactions (AME × CP, AME × Lys, CP × Lys, and AME × CP × Lys) were included in the model as fixed effects. Results are presented as means and standard error of the mean (SEM), with statistical significance set at P < 0.05. When ANOVA indicated main effects or interactions (P < 0.05), Tukey’s HSD test was applied to estimated marginal means (EMMs). Main effects were compared using marginal means averaged over other factors, while interactions were analyzed using simple-effect comparisons within interaction slices. If a three-way interaction was detected, the interpretation prioritized the three-way interaction and the main effects of each factor. Quadratic polynomial models for ADG and FCR were constructed using the custom design module in Design-Expert 13 software (Version 13, Stat-Ease Inc., Minneapolis, MN, USA) based on pen-level data. The model equation was as follows:

$$\mathbf{Y}={\mathit{\beta }}_{\mathbf{0}}+{\mathit{\beta }}_{\mathbf{1}}{\mathbf{X}}_{\mathbf{1}}+{\mathit{\beta }}_{\mathbf{2}}{\mathbf{X}}_{\mathbf{2}}+{\mathit{\beta }}_{\mathbf{3}}{\mathbf{X}}_{\mathbf{3}}+{\mathit{\beta }}_{\mathbf{12}}{\mathbf{X}}_{\mathbf{1}}{\mathbf{X}}_{\mathbf{2}}+{\mathit{\beta }}_{\mathbf{13}}{\mathbf{X}}_{\mathbf{1}}{\mathbf{X}}_{\mathbf{3}}+{\mathit{\beta }}_{\mathbf{23}}{\mathbf{X}}_{\mathbf{2}}{\mathbf{X}}_{\mathbf{3}}+{\mathit{\beta }}_{\mathbf{11}}{{\mathbf{X}}_{\mathbf{1}}}^{\mathbf{2}}+{\mathit{\beta }}_{\mathbf{22}}{{\mathbf{X}}_{\mathbf{2}}}^{\mathbf{2}}+{\mathit{\beta }}_{\mathbf{33}}{{\mathbf{X}}_{\mathbf{3}}}^{\mathbf{2}}$$

where Y is the response of interest (ADG or FCR), X₁ is AME, X₂ is CP, X₃ is Lys, β₀ is the intercept, and β₁, β₂, β₃, β₁₂, β₁₃, β₂₃, β₁₁, β₂₂ and β₃₃ are the coefficients estimated by the model.

Results

Growth performance

As shown in Table 2, dietary AME levels affected the ADFI (P < 0.0001) and FCR (P < 0.0001) of starter Pekin ducks. Both ADFI and FCR decreased as dietary AME levels increased. Meanwhile, dietary CP levels affected on ADG (P = 0.0011), ADFI (P < 0.0001), and FCR (P = 0.0006). Compared with the 17 and 18.5% CP groups, the 20% CP group reduced ADG and ADFI. Furthermore, FCR decreased in the 18.5 and 20% CP groups compared to the 17% CP group. Dietary Lys levels also influenced ADG (P < 0.0001), ADFI (P < 0.0001), and FCR (P < 0.0001). Specifically, increasing dietary Lys levels increased ADG and reduced FCR. Compared with the 0.8% Lys group, ADFI increased in the 1.05 and 1.3% Lys groups.

Table 2.

Effects of dietary AME, CP, and Lys levels on growth performance of starter Pekin ducks.

AME (kcal/kg)	CP (%)	Lys (%)	ADG (g/d)	ADFI (g/d)	FCR (g/g)
2,750	17	0.8	23.52	38.06	1.62
		1.05	38.38	55.10	1.44
		1.3	39.83	57.41	1.44
	18.5	0.8	22.80	36.64	1.61
		1.05	41.89	60.14	1.44
		1.3	45.52	63.20	1.39
	20	0.8	19.87	32.56	1.65
		1.05	38.54	54.69	1.42
		1.3	42.07	59.17	1.41
2,900	17	0.8	24.82	39.15	1.58
		1.05	41.91	58.05	1.39
		1.3	42.98	59.39	1.38
	18.5	0.8	22.53	35.83	1.59
		1.05	42.08	58.04	1.38
		1.3	41.87	57.75	1.38
	20	0.8	19.13	31.66	1.66
		1.05	40.12	54.14	1.35
		1.3	42.13	56.78	1.35
3,050	17	0.8	24.53	38.68	1.58
		1.05	39.48	53.20	1.35
		1.3	42.12	55.86	1.33
	18.5	0.8	21.00	32.75	1.56
		1.05	42.63	56.38	1.32
		1.3	43.51	56.17	1.29
	20	0.8	19.59	30.26	1.55
		1.05	41.13	54.22	1.32
		1.3	43.38	55.45	1.28
3,200	17	0.8	25.58	37.87	1.48
		1.05	40.23	51.24	1.28
		1.3	40.36	51.26	1.27
	18.5	0.8	22.82	33.84	1.48
		1.05	41.89	53.18	1.27
		1.3	42.67	52.98	1.24
	20	0.8	20.41	30.25	1.48
		1.05	41.04	51.61	1.26
		1.3	42.03	51.87	1.23
SEM			1.19	1.55	0.01
AME
2,750			34.71	50.77a	1.49a
2,900			35.29	50.09a	1.45b
3,050			35.26	48.11b	1.40c
3,200			35.22	46.01c	1.33d
CP
17			35.31a	49.60a	1.43a
18.5			35.93a	49.74a	1.41b
20			34.12b	46.89b	1.41b
Lys
0.8			22.22c	34.79b	1.57a
1.05			40.78b	55.00a	1.35b
1.3			42.37a	56.44a	1.33c
P-value
AME			0.6990	<0.0001	<0.0001
CP			0.0011	<0.0001	0.0006
Lys			<0.0001	<0.0001	<0.0001
AME × CP			0.1424	0.1479	0.0441
AME × Lys			0.4101	0.0447	0.0007
CP × Lys			<0.0001	<0.0001	<0.0001
AME × CP × Lys			0.9666	0.9792	0.0077

Presented as mean and standard error of the mean (SEM) (n = 4).

Within a column, means with different superscripts are different (P < 0.05).

Abbreviations: AME, apparent metabolizable energy; CP, crude protein; Lys, lysine; ADG, average daily gain; ADFI, average daily feed intake; FCR, feed conversion ratio.

Interactions between dietary CP and Lys were observed for the ADG (P < 0.0001) and ADFI (P < 0.0001) of starter Pekin ducks (Table 2). Specifically, when the dietary Lys level was at 0.8%, both ADG and ADFI decreased with increasing CP levels (Fig. 1A and B). However, when dietary Lys was at 1.05 or 1.3%, increasing dietary CP improved ADG. At the 1.3% Lys level, increasing dietary CP did not effect on ADFI. Simultaneously, an interaction between dietary AME and Lys levels was observed for ADFI (P = 0.0447). When dietary Lys was 0.8%, AME levels did not influence ADFI, whereas at 1.05 or 1.3% Lys, ADFI decreased as AME levels increased (Fig. 1C).

Fig. 1.

Effects of different dietary apparent metabolizable energy (AME), crude protein (CP), and lysine (Lys) levels on growth performance. (A) Effects of CP levels on average daily gain (ADG) under different Lys levels; (B) Effects of CP levels on average daily feed intake (ADFI) under different Lys levels; (C) Effects of AME levels on ADFI under different Lys levels; (D) Effects of Lys levels on feed conversion ratio (FCR) at AME levels of 3,050 and 3,200 kcal/kg; (E) Effects of CP levels on FCR at AME levels of 3,050 and 3,200 kcal/kg; (F) Effects of CP levels on FCR under low AME (2,750 and 2,900 kcal/kg) and different Lys levels; (G) Effects of CP levels on FCR under high AME (3,050 and 3,200 kcal/kg) and different Lys levels. Data are presented as mean ± standard error of the mean (SEM). Different letters indicate differences (P < 0.05). Panels D and E show main effects of AME levels without interactions.

As shown in Table 2, a three-way interaction among dietary AME, CP, and Lys was detected for the FCR of starter Pekin ducks (P = 0.0077). The effect of CP on FCR depended on both AME and Lys levels (Fig. 1F-G). At low AME (2,750–2,900 kcal/kg) combined with Lys deficiency (0.80%), increasing CP worsened FCR (Fig. 1F). At low AME but adequate Lys (1.30%), increasing CP improved FCR. At high AME (3,050–3,200 kcal/kg), CP and Lys acted independently to reduce FCR, with no interaction (Fig. 1D-E).

Slaughter performance

As shown in Table 3, dietary AME affected the breast muscle yield (P = 0.0076) and abdominal fat percentage (P < 0.0001) of starter Pekin ducks, but did not effect leg muscle yield. Compared with the 2,750 and 3,050 kcal/kg AME groups, the breast muscle yield of the 3,200 kcal/kg AME group was lower. The abdominal fat percentage increased as dietary AME levels rose. Dietary CP influenced abdominal fat percentage (P < 0.0001), but did not affect breast or leg muscle yields. Compared to the 17% CP group, the abdominal fat percentage of the 18.5 and 20% CP groups was lower. Dietary Lys affected breast muscle yield (P < 0.0001), leg muscle yield (P < 0.0001), and abdominal fat percentage (P < 0.0001). Compared with the 0.8% Lys group, the 1.05 and 1.3% Lys groups had greater breast and leg muscle yields and lower abdominal fat percentage.

Table 3.

Effects of dietary AME, CP, and Lys levels on slaughter performance of starter Pekin ducks.

AME (kcal/kg)	CP (%)	Lys (%)	Breast muscle yield (%)	Leg muscle yield (%)	Abdominal fat (%)
2,750	17	0.8	0.89	9.06	0.29
		1.05	1.66	10.76	0.32
		1.3	1.42	9.24	0.22
	18.5	0.8	1.00	9.42	0.28
		1.05	1.56	9.97	0.24
		1.3	1.68	10.13	0.26
	20	0.8	0.89	9.71	0.21
		1.05	1.47	9.87	0.24
		1.3	1.72	10.15	0.20
2,900	17	0.8	0.99	9.53	0.41
		1.05	1.63	10.07	0.30
		1.3	1.67	10.42	0.32
	18.5	0.8	0.91	9.49	0.21
		1.05	1.44	10.68	0.13
		1.3	1.47	10.68	0.15
	20	0.8	0.86	9.33	0.26
		1.05	1.39	10.09	0.20
		1.3	1.66	9.37	0.22
3,050	17	0.8	1.00	9.74	0.52
		1.05	1.53	9.80	0.40
		1.3	1.55	9.90	0.30
	18.5	0.8	0.97	9.09	0.47
		1.05	1.57	10.32	0.37
		1.3	1.76	10.30	0.34
	20	0.8	0.98	9.46	0.27
		1.05	1.49	10.80	0.27
		1.3	1.66	10.03	0.28
3,200	17	0.8	0.89	9.25	0.62
		1.05	1.49	10.06	0.43
		1.3	1.42	9.61	0.39
	18.5	0.8	1.00	9.02	0.63
		1.05	1.47	10.64	0.40
		1.3	1.51	10.74	0.26
	20	0.8	0.94	9.98	0.42
		1.05	1.46	10.39	0.35
		1.3	1.32	10.50	0.31
SEM			0.07	0.47	0.04
AME
2,750			1.36a	9.81	0.25c
2,900			1.33ab	9.96	0.24c
3,050			1.39a	9.94	0.35b
3,200			1.28b	10.02	0.42a
CP
17			1.34	9.79	0.38a
18.5			1.36	10.04	0.31b
20			1.32	9.97	0.27b
Lys
0.8			0.94b	9.42b	0.38a
1.05			1.51a	10.29a	0.30b
1.3			1.57a	10.09a	0.27b
P-value
AME			0.0076	0.8192	<0.0001
CP			0.3987	0.3954	<0.0001
Lys			<0.0001	<0.0001	<0.0001
AME × CP			0.0546	0.4936	0.0041
AME × Lys			0.0869	0.9945	0.0017
CP × Lys			0.0832	0.4046	0.0670
AME × CP × Lys			0.2656	0.6435	0.5976

Presented as mean and standard error of the mean (SEM) (n = 4).

Within a column, means with different superscripts are different (P < 0.05).

Abbreviations: AME, apparent metabolizable energy; CP, crude protein; Lys, lysine.

Interactions between dietary AME and CP levels were observed for the abdominal fat percentage of starter Pekin ducks (P = 0.0041; Table 3). Specifically, when dietary AME was 2,750 kcal/kg, CP levels did not effect abdominal fat percentage. However, at AME levels of 2,900, 3,050, or 3,200 kcal/kg, increasing dietary CP reduced the abdominal fat percentage (Fig. 2A). Simultaneously, dietary AME and Lys interactions were observed for abdominal fat percentage (P = 0.0017; Table 3). When dietary AME was 2,750 or 2,900 kcal/kg, Lys levels did not influence abdominal fat percentage. In contrast, at AME levels of 3,050 or 3,200 kcal/kg, increasing dietary Lys reduced abdominal fat percentage (Fig. 2B).

Fig. 2.

Effects of different dietary apparent metabolizable energy (AME), crude protein (CP), and lysine (Lys) levels on abdominal fat percentage. (A) Effects of CP levels on abdominal fat percentage under different AME levels; (B) Effects of Lys levels on abdominal fat percentage under different AME levels. Data are presented as mean ± standard error of the mean (SEM). Different letters indicate differences (P < 0.05).

Estimation of AME, CP, and Lys requirements

Quadratic polynomial response surface models for ADG and FCR were developed using Design-Expert 13 software based on pen-level data. For both response variables, the overall regression models were significant (P < 0.0001), indicating that the fitted models adequately described the relationships between dietary AME, CP, Lys, and performance responses. Residual diagnostics, including normal probability plots, residuals versus predicted values, residuals versus run order, and Cook’s distance analysis, indicated that the assumptions of normality, homoscedasticity, and independence were adequately satisfied, with no influential outliers detected for the regression models of ADG (Figure S1) and FCR (Figure S2), respectively.

For ADG, the model showed good explanatory power, with an adjusted R² of 0.9371 and a low RMSE of 2.39, and no significant lack-of-fit was detected (P = 0.3768). These results indicate a reliable model fit with strong predictive ability. The predicted optimal solution for ADG was located within the experimental design space.

For FCR, the model also showed high explanatory power, with an adjusted R² of 0.9630 and an RMSE of 0.024, indicating strong goodness of fit. However, a significant lack-of-fit was observed (P = 0.0003), suggesting unexplained variation in the model. The predicted optimum for FCR was located near the boundary of the experimental design space.

The quadratic polynomial equation fitted for ADG based on the uncoded (actual) independent variables is as follows:

$$\mathbf{ADG}=-\mathbf{323}.\mathbf{87279}+\mathbf{0}.\mathbf{042851}\times \mathbf{AME}+\mathbf{15}.\mathbf{31964}\times \mathbf{CP}+\mathbf{268}.\mathbf{74862}\times \mathbf{Lys}+\mathbf{0}.\mathbf{000130}\times \mathbf{AME}\times \mathbf{CP}-\mathbf{0}.\mathbf{005245}\times \mathbf{AME}\times \mathbf{Lys}+\mathbf{3}.\mathbf{92096}\times \mathbf{CP}\times \mathbf{Lys}-\mathbf{6}.\mathbf{50829}\times \mathbf{1}{\mathbf{0}}^{-\mathbf{6}}\times \mathbf{AM}{\mathbf{E}}^{\mathbf{2}}-\mathbf{0}.\mathbf{546347}\times \mathbf{C}{\mathbf{P}}^{\mathbf{2}}-\mathbf{135}.\mathbf{87020}\times \mathbf{Ly}{\mathbf{s}}^{\mathbf{2}}(\mathbf{adjusted}{\mathbf{R}}^{\mathbf{2}}=\mathbf{0}.\mathbf{9371})$$

Fig. 3 illustrates the response surface and contour plots for ADG, which exhibit a peak point within the experimental range. According to the model, the maximum predicted ADG of 44.79 g/d for starter Pekin ducks is achieved when the dietary AME, CP, and Lys levels are 2,995 kcal/kg, 18.65%, and 1.20%, respectively.

Fig. 3.

Response surface and contour plots illustrating the relationship between ADG (g/d) and dietary AME (kcal/kg), CP (%), and Lys (%).

The quadratic polynomial equation fitted for FCR based on the uncoded (actual) independent variables is as follows:

$$\mathbf{FCR}=\mathbf{1}.\mathbf{45206}+\mathbf{0}.\mathbf{001693}\times \mathbf{AME}-\mathbf{0}.\mathbf{036502}\times \mathbf{CP}-\mathbf{2}.\mathbf{79775}\times \mathbf{Lys}-\mathbf{0}.\mathbf{000013}\times \mathbf{AME}\times \mathbf{CP}-\mathbf{0}.\mathbf{000124}\times \mathbf{AME}\times \mathbf{Lys}-\mathbf{0}.\mathbf{036932}\times \mathbf{CP}\times \mathbf{Lys}-\mathbf{2}.\mathbf{80820}\times \mathbf{1}{\mathbf{0}}^{-\mathbf{7}}\times \mathbf{AM}{\mathbf{E}}^{\mathbf{2}}+\mathbf{0}.\mathbf{002934}\times \mathbf{C}{\mathbf{P}}^{\mathbf{2}}+\mathbf{1}.\mathbf{60839}\times \mathbf{Ly}{\mathbf{s}}^{\mathbf{2}}(\mathbf{adjusted}{\mathbf{R}}^{\mathbf{2}}=\mathbf{0}.\mathbf{9630})$$

Fig. 4 presents the response surface and contour plots for the FCR, which exhibit a minimum point within the experimental design. The model predicts that the minimum FCR of 1.22 g/g for starter Pekin ducks is achieved when the dietary AME, CP, and Lys levels are 3,199 kcal/kg, 19.58%, and 1.15%, respectively.

Fig. 4.

Response surface and contour plots illustrating the relationship between FCR (g/g) and dietary AME (kcal/kg), CP (%), and Lys (%).

Discussion

Dietary levels of AME, CP, and Lys are the core nutritional factors influencing poultry growth performance (Li et al., 2024). In the present study, a 4 × 3 × 3 factorial experimental design was employed to systematically evaluate the effects of dietary AME, CP, and Lys levels, as well as their interactions, on the growth and slaughter performance of starter Pekin ducks.

In previous studies in ducks, Wen et al. (2017) and Wickramasuriya et al. (2016a) reported that appropriate Lys supplementation improved weight gain and feed intake in Pekin ducks and Korean native ducks, respectively. Similar phenomena have also been found in broilers by Tran et al. (2021) and Sterling et al. (2006). Consistent with these findings, increasing dietary Lys from 0.8 to 1.05% improved ADG and ADFI in starter Pekin ducks, indicating that 0.8% Lys is insufficient to meet their nutritional requirements. Generally, increasing dietary CP exerts a positive effect on poultry growth performance, as confirmed by numerous studies (Sterling et al., 2003; Min et al., 2007; Wang et al., 2020; Shen et al., 2025), which is consistent with our findings under adequate Lys levels. However, when the Lys level was 0.8%, increasing dietary CP conversely decreased the ADG and ADFI of starter Pekin ducks. This may be attributed to the fact that Lys deficiency restricts the efficiency of protein synthesis (Hasan et al., 2020), leading to the ineffective utilization and subsequent catabolism of other amino acids (Macelline et al., 2025). Furthermore, amino acid imbalances may inhibit feed intake regulation mechanisms, thereby reducing total intake (Bjordal et al., 2014). Previous research has indicated that the requirement for an individual limiting amino acid decreases correspondingly under low protein conditions (Hurwitz et al., 1998), whereas increasing the protein level may further exacerbate the imbalance. Therefore, under conditions of Lys deficiency, simply increasing the protein level not only fails to improve growth performance but may also lead to detrimental effects.

Energy is the primary factor influencing feed intake in poultry. Birds adjust their feed consumption based on dietary AME levels to ensure adequate energy intake for maintaining normal physiological activities (Wang et al., 2025). High energy diets typically suppress appetite, whereas low energy diets stimulate feed intake (Hu et al., 2019; Hong et al., 2022). The present study demonstrated that ADFI decreased with increasing dietary AME levels in starter Pekin ducks. This finding is consistent with Wu et al. (2019), who observed that increasing dietary AME from 10.82 MJ/kg to 12.95 MJ/kg reduced the ADFI of growing Pekin ducks. Furthermore, it was found that Lys deficiency (0.8%) reduced feed intake, and the regulatory effect of dietary AME on ADFI was modulated by Lys levels. Specifically, the phenomenon of ADFI decreasing with increased AME levels only occurred when Lys was at higher levels (1.05 and 1.3%). This may be because Lys deficiency activates the GCN2 pathway, which inhibits the consumption of amino acid-imbalanced diets (Bjordal et al., 2014; Shen and Zhu, 2017). Under Lys deficient conditions, feed intake is already at a low level, thereby limiting the regulatory capacity of dietary AME. Once dietary Lys was adequate (1.05 and 1.3%), birds recovered their normal physiological status, allowing AME to exert its regulatory control over feed intake via the activation or inhibition of the AMPK pathway (Hardie et al., 2012; Hu et al., 2019).

Available literature indicates that increasing either dietary AME or CP levels can effectively reduce the FCR of Pekin ducks (Zeng et al., 2015). Wen et al. (2017) and Wickramasuriya et al. (2016a) further pointed out that increasing dietary Lys can lower the FCR of Pekin ducks. Consistent with these findings, the present study demonstrated that increasing dietary AME, CP, and Lys levels all contributed to a reduction in the FCR of starter Pekin ducks. Moreover, a three-way interaction among dietary AME, CP, and Lys was observed for FCR. Under conditions of low AME and low Lys, increasing CP levels conversely raised the FCR. This may be attributed to the dual restriction of energy and the limiting amino acid, which hinders protein deposition. Consequently, excess amino acids that cannot be utilized for body protein synthesis are deaminated and catabolized for energy (Macelline et al., 2025), thereby increasing energy expenditure and reducing utilization efficiency (Bender, 2012). Research has shown that amino acid imbalances lead to the degradation of surplus amino acids, incurring additional metabolic costs (Liu et al., 2021), and under energy deficient conditions, amino acids can also be deaminated and enter oxidative pathways to provide energy (Torres et al., 2023). Conversely, under low energy but Lys adequate conditions, increasing CP levels maintained or reduced FCR, suggesting that protein can be effectively utilized for growth and deposition when the requirement for the limiting amino acid is met (Zhao et al., 2019). Similar three-way interactions were reported in broiler research, where interactions among CP, energy, and amino acids influenced not only FCR, but also weight gain, feed intake, and breast muscle yield (Musigwa et al., 2025).

The breast and leg muscles are the primary edible portions of ducks, with higher yields indicating greater economic benefits (Wen et al., 2017). Lys has been reported to enhance breast meat yield in both broilers (Tang et al., 2007) and Pekin ducks (Xie et al., 2009). Furthermore, Wen et al. (2017) observed that supplemental Lys improved the leg muscle yields in Pekin ducks. Consistent with these findings, the present study demonstrated that increasing dietary Lys from 0.8 to 1.05% increased both breast and leg meat yields in starter Pekin ducks, while further increasing it to 1.3% provided no additional improvement. This suggests that a 0.8% Lys level is insufficient to meet the muscle tissue growth and development requirements of starter Pekin ducks. Notably, a study on broilers revealed that birds fed a Lys deficient starter diet failed to achieve breast meat yields comparable to the Lys adequate group at market age, even when Lys levels were restored during the grower phase (Kidd et al., 1998). This indicates that Lys deficiency during the starter period may exert irreversible negative impacts on muscle development.

Excessive abdominal fat is not only regarded as a production waste, but also as a sign of reduced feed utilization efficiency in poultry. Dietary AME is a critical factor influencing abdominal fat deposition. Studies by Maharjan et al. (2021) and Ko et al. (2023) on broilers demonstrated that increasing dietary AME increased abdominal fat deposition in broilers. Consistent with these findings, the present study observed that the abdominal fat percentage of Pekin ducks increased with increasing dietary AME. Furthermore, interactions between AME and CP, as well as AME and Lys, were found to influence the abdominal fat percentage of starter Pekin ducks. Specifically, at higher AME levels (3,050 and 3,200 kcal/kg), increasing dietary CP or Lys reduced abdominal fat percentage, whereas no effects were observed at the low AME level (2,750 kcal/kg). This is similar to the findings of Griffiths et al. (1977), which showed that decreasing the dietary energy-to-protein ratio reduced the proportion of abdominal fat in broilers. Research by Huang et al. (2013) also indicated that diets with a low protein-to-carbohydrate ratio increase energy intake as well as white and brown adipose tissue mass in mice. It has been found that high protein intake down regulates the expression of enzymes involved in lipogenesis and reduces the utilization efficiency of amino acid derived carbon skeletons in lipogenic pathways, thereby decreasing de novo fatty acid synthesis (Chaumontet et al., 2015). Additionally, the thermic effect of food during the digestion, absorption, and metabolism of protein is substantially higher than that of carbohydrates and fats (Westerterp, 2004), which further increases energy expenditure and subsequently minimizes energy surplus and fat deposition. Consequently, higher dietary CP or Lys levels can, to some extent, alleviate the excessive abdominal fat deposition induced by high energy diets.

Precise supply of nutrients during the starter period is critical for optimizing growth performance and reducing feeding costs. Traditionally, nutritional requirements have predominantly been predicted using single factor dose response experiments. For instance, Xie et al. (2009) employed broken-line models to predict that the Lys levels required for optimal weight gain and FCR in Pekin ducks (7–21 d) were 0.84 and 0.90%, respectively. Similarly, Wickramasuriya et al. (2016b) used linear and quadratic plateau models to estimate that the AME levels required for maximum weight gain and minimum FCR in Korean native ducks (1–21 d) were 2,953 and 2,950 kcal/kg, respectively. Compared to single factor models, multifactorial models better capture the interactions among multiple nutrients and align more closely with practical production applications. For example, Luo et al. (2025) utilized a Box-Behnken design to construct a response surface model, determining that the net energy, standardized ileal digestible Lys, and amylose-to-amylopectin ratio required for maximum body weight in 20-D-old broilers were 2,303 kcal/kg, 1.24%, and 0.22, respectively. In the present study, a 4 × 3 × 3 factorial design was used to construct quadratic polynomial models for estimating the nutritional requirements of starter Pekin ducks. Our results indicate that to achieve maximum ADG, the dietary AME, CP, and Lys levels should be 2,995 kcal/kg, 18.65%, and 1.20%, respectively. To achieve the minimum FCR, these levels are 3,199 kcal/kg, 19.58%, and 1.15%, respectively. The predicted AME level was close to the upper boundary of the experimental range (3,200 kcal/kg), suggesting that the response may not have fully reached a plateau and that higher AME levels could potentially further influence FCR. Although a significant lack-of-fit was detected for the FCR model, the overall model performance remained acceptable, as indicated by high R² values and consistent biological trends within the experimental range. Across both ADG and FCR models, analyzed values of dietary CP were overall slightly higher than the calculated values, a consistent gradient was maintained between groups. Furthermore, the model's fitting performance for ADG and FCR (R² > 0.94) demonstrated that multifactorial regression analysis has certain practical utility in evaluating nutrient synergy effects, providing data references for formulating diets of starter Pekin ducks.

The estimated CP requirement in the present study was consistent with the value reported by Xie et al. (2017). In contrast, the optimal AME and Lys levels were slightly higher than those reported by Xie et al. (2010); Wickramasuriya et al. (2016b), and Wen et al. (2017). Such differences may be related to variations in duck genotype, diet composition, experimental conditions, and the response criteria used for requirement estimation. Furthermore, the response surface approach used in the present study accounted for interactions among AME, CP, and Lys, which may have resulted in different nutrient recommendations compared with those derived from single-factor studies.

Several limitations of this study should be considered when interpreting the results. Sex was not considered as an experimental factor in the present study. Therefore, potential sex-related differences in nutrient requirements could not be evaluated and warrant further investigation. Moreover, no additional validation trials were conducted; therefore, the models are predictive and should be confirmed experimentally.

Conclusion

In conclusion, dietary AME, CP, Lys, and their interactions influenced the growth and slaughter performance of starter Pekin ducks. Lys deficiency severely impaired production performance. Notably, increasing dietary CP levels under Lys deficient conditions further reduced feed intake and weight gain, which may be attributed to disrupted amino acid balance. Increasing dietary Lys levels effectively reduced FCR and promoted breast and leg muscle development, while simultaneously alleviating excessive abdominal fat deposition induced by high energy diets. The effect of dietary AME on feed intake was related to the level of Lys. Under conditions of Lys deficiency, changing the AME level failed to regulate feed intake. Furthermore, the three-way interaction among AME, CP, and Lys affected FCR. Under low AME and low Lys conditions, increasing the CP level resulted in a higher FCR. Regression analysis indicated that the optimal dietary levels to achieve maximum ADG were 2,995 kcal/kg AME, 18.65% CP, and 1.20% Lys, while the optimal levels for minimum FCR were estimated at 3,199 kcal/kg AME, 19.58% CP, and 1.15% Lys.

Authorship contribution

Yalong Deng designed and conducted the study, coordinated the animal trial, collected and analyzed the data, interpreted the results, and drafted the manuscript. Jing Tang contributed to the experimental design, supervised the study, assisted with data analysis and interpretation, and critically revised the manuscript. Xianglong Yun, Xin Wen, Yongyan Jin, Lipeng Yuan, Qingyi Wu, Zhongjian Shen, Deying Ma, Zhengkui Zhou, and Ming Xie contributed to the conduct of the study, sample collection, data organization and interpretation, participated in the discussion of the results, and critically reviewed the manuscript. Shuisheng Hou conceived and supervised the project, contributed to the study design and interpretation of the findings, secured funding, and critically revised the manuscript. All authors contributed to the article, reviewed and approved the final version of the manuscript, and agreed to be accountable for all aspects of the work.

Disclosures

The authors declare that there is no conflict of interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.