The leaves and tender tops of fresh ramie (Boehmeria nivea cv. Qingsizhu No. 1) were cut at about 80 cm height and immediately dried at 60°C for 4 days in a heat drier room. The weight ratio of leaf to stem in dry ramie was 3.37. Then, the dried stems and leaves were crushed to powder using a grinder equipped with a 1.5 mm sieve, and kept in a well-closed and light-resistant room.

Three hundred and twelve 21-day-old female Linwu ducks, free of infectious disease, were obtained from Hunan Shunhua Duck Industrial Development Company, and transferred to the laboratory of the Institute of Bast Fiber Crops, Chinese Academy of Agricultural Sciences. The ducks were supplied ad libitum access to feed and water throughout the trial period. After a 1-week adaptation period fed with the control diet, all the ducks were individually weighed and equally divided into 4 groups, meeting the purpose that the average initial weights among groups were not significant different. The ducks in each group (78 Linwu ducks) were further subdivided into six 120 cm × 120 cm cages (13 ducks/cage). Group 1 received the control diet which was based on corn, soybean meal, wheat bran, wheat flour, and rice husk. Group 2, 3, and 4 received a control diet in which corn, soybean meal, wheat bran, wheat flour, and rice husk were partly replaced by ramie powder to reach proportions of 3, 6, or 12%, respectively. All diets were formulated with similar levels of nutrient and to meet recommendations of Nutrient Requirements of Meat-type Duck (China, NY/T 2122-2012), as shown in Table 1. The feeding period was 42 days.

Body weight of Linwu duck was individually measured at the beginning (28-day-old) and the end of the trial (70-day-old). Feed intake per cage was recorded daily. The average daily feed intake (ADFI), average daily body weight gain (ADG), and feed/gain ratios (F/G) were calculated for each cage.

On the last day of the trial, twelve ducks in each group (two ducks in each cage) with body weights close to the mean were selected after 12-h fasting. Blood samples were collected via wing vein with 10-mL heparin-free tubes, and centrifuged at 3,000 × g for 15 min at 4°C. The obtained serum samples were stored at −20°C until analysis. The ducks were then immediately slaughtered by cervical dislocation. The carcasses were bled by hanging for 5 min and scalded in water (65°C) for approximately 1-min for feathers plucking. The weights of carcasses were recorded as dressed weight. After carefully excising the gastrointestinal tract and organs, the weights of half-eviscerated, eviscerated, abdominal fat, leg muscle, breast muscle, and lean meat (total legs and breasts muscle) were recorded. Then, a 1-cm-thick sample of leg muscle and breast muscle was rapidly taken from each duck and frozen in liquid nitrogen and stored at −80°C until RNA extraction. Finally, the dressed percentage (PD), percentage of half-eviscerated yield (PHEY), percentage of eviscerated yield (PEY), percentage of breast muscle (PB), percentage of leg muscle (PL), percentage of lean meat (PLM), and percentage of abdominal fat (PAF) were calculated, respectively, according to the regulations and requirements of Performance Ferms and Measurement for Poultry (China, NY/T 823-2004).

The meat qualities of leg and breast muscles were measured as follow: the meat colors (45-min and 24-h postmortem) were determined as the L*, a*, and b*, which were the indicators of lightness, redness and yellowness, respectively, with a colorimeter (CR-400; Minolta Camera Co., Osaka, Japan). The pH values (45-min and 24-h postmortem) were measured with a pH meter (Model 340, Mettler-Toledo GmbH, Schwerzenbach, Switzerland) by the method described previously (Choi et al., 2016). The drip losses at 24-h postmortem, the cooking losses at 45-min and 24-h postmortem were assayed according to the method described previously (Qiao et al., 2017), and the shear forces were further measured with the digital tenderness meter (C-LM3B, Northeast Agricultural University, Harbin, China) after measuring the cooking loss, to evaluate the tenderness of meat (Tang et al., 2009).

The serum levels of glutathione (GSH), malonaldehyde (MDA) and the activities of superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPX), and total antioxidant capacity (T-AOC) were determined by the commercial kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) and an automated fluorescence instrument (Thermo Fisher Scientific, Waltham, MA, United States) following the corresponding procedures. And, the serum level of 8-hydroxy-2′-deoxyguanosine (8-OHdG) was quantified by a specific ELISA kit (Abcam, Cambridge, United Kingdom).

Total RNA from the leg muscle and breast muscle were isolated using Trizol reagent (TaKaRa, Tokyo, Japan), and then treated with DNase I (Thermo Fisher Scientific Inc., Waltham, MA, United States). The cDNA were synthesized from 1 μg of RNA with a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific Inc., Waltham, MA, United States) according to the manufacturer’s instructions. Based on the cloned complete sequences¹ of nuclear factor erythroid-2-related factor 2/erythroid-derived CNC homology factor (Nrf2/ECH), catalase (CAT), superoxide dismutase-1 (SOD1), glutathione peroxidase-1 (GPX1), and β-actin from Anas platyrhynchos, primer pairs were designed with Primer 5.0 for quantitative real-time PCR (Table 2). The β-actin gene was used as the housekeeping gene. All primers were synthesized and purified by Sangon Biotech Co. Ltd. (Shanghai, China). Reaction volume of 20 μL mixture contained 10 μL Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, United States), 1 μL cDNA template, 1 μL of the upstream and downstream primers for each targeting gene, and 7 μL sterilized deionized water. The amplification parameters for all the genes of the thermocycler (CFX Connect, Bio-Rad, Inc., Hercules, CA, United States) were a preheat period of 3 min at 95°C followed by 45 cycles of 95°C for 10 s and 55°C for 20 s, and a melting curve ramping from 65 to 95°C with an increasing temperature of 0.5°C. All sample analyses were carried out in triplicate and the average values were indexed. The target gene expressions were normalized to that of the selected reference gene, and the relative gene expressions were calculated using 2^–ΔΔCt method (Livak and Schmittgen, 2001). The mRNA levels were expressed as the fold change relative to the mean values of the control group, which was arbitrarily defined as 1.0.

Statistical analysis of data was conducted using Statistical Package for the Social Sciences 19.0 (IBM, Armonk, NY, United States). One-way ANOVA model was performed in the present research. Replicate was used as the experimental unit (As for the results of the biomarker analysis, it means that the results of two birds in each cage are summarized as an average value). Results were presented as means and pooled standard errors of the means (SEM). Differences among means of all groups were considered significant at P < 0.05. The P values between 0.05 and 0.10 were considered as a trend. Duncan’s multiple range were further used when the main effect was significant. And the orthogonal polynomial contrasts were used to determine linear and quadratic responses of measured parameters of Linwu ducks to ramie powder levels.

Growth performances of ducks in different treatment groups were shown in Table 3. In comparison with the control group, the groups fed diets supplemented with ramie level at 6 and 12% had a significant increase in the average final weight, and ADG (P < 0.05) and had a trend in increase the ADFI (P = 0.063). Moreover, the greatest average final weight, ADG, and ADFI were achieved for ducks fed with the 6% ramie supplemented diet. Positive linear relationships were found between ramie dose and the average final weight (P = 0.009), ADG (P = 0.006), and ADFI (P = 0.013). However, there were no significant differences (P > 0.05) in F/G ratio among groups.

Carcass characteristics were shown in Table 4. There were no significant differences (P > 0.05) in carcass characteristics among groups.

Meat qualities of the Linwu ducks were shown in Tables 5, 6. In breast muscles, no effect of dietary ramie supplementation was detected on the shear force, 24 h pH value, 45 min and 24 h meat color or cooking loss. However, there were trends in increase in the pH value at 45 min (P = 0.081) and decrease in the drip loss (P = 0.096) in the groups treated with ramie. In addition, quadratic effects (P < 0.05) on drip loss and pH value at 45 min were observed as the dietary ramie level increased from 0 to 12%, and the lowest drip loss value and the highest 45 min pH value were noted, respectively, in ducks fed with 6 and 3% ramie.

In leg muscles, compared with the control group and 12% ramie treated group, the 6% ramie treated group had a significant decrease in the cooking loss at 45 min (P = 0.043). Compared with the 12% ramie treated group, the 6% ramie treated group had a significant increase in the pH value at 45 min (P = 0.048). Quadratic effects (P < 0.05) for 45 min cooking loss and pH value were observed as the dietary ramie level increased from 0 to 12%. However, there were no significant differences (P > 0.05) on other characteristics of leg muscles quality among groups.

Levels of serum antioxidant biomarkers were shown in Table 7. In comparison with the control group, the groups supplemented with 6% ramie had a significant increase in the content of GSH and the activity of SOD (P < 0.05). In addition, there were increasing trends in the levels of GPX (P = 0.082) and T-AOC (P = 0.092) in the groups supplemented with ramie. Moreover, positive linear relationships were observed between ramie dose and serum levels of GPX and GSH (P < 0.05). Besides, significant quadratic relationships were observed between ramie dose and serum levels of SOD, and T-AOC (P < 0.05).

The mRNA expression levels of muscular antioxidant-related genes were shown in Table 8. In breast muscles, compared with the control group, the group with ramie level at 6% had significant increase in the GPX1 mRNA expression level (P < 0.05). A quadratic effect was observed between the dietary ramie level and for GPX1 mRNA expression level (P = 0.012).

In leg muscles, compared with the control group and the 12% ramie treated group, the 6% ramie treated group showed a significant increase in the SOD1 mRNA expression level (P = 0.048). There was an increasing trend in the CAT mRNA expression level in the ramie treated groups (P = 0.098). In addition, quadratic effects were observed between dietary ramie level and SOD1 and CAT mRNA expression level (P < 0.05). However, dietary ramie supplementation did not influence the mRNA expression levels of Nrf2/ECH in either breast muscles or leg muscles of the ducks.