Effects of legume-diet and sex of ducks on the growth performance, physicochemical traits of meat and fatty acid composition in fat

Duck meat enjoys growing popularity among consumers. Alternative protein sources to soybean are being investigated to eliminate genetically modified components from the poultry’ diet. The aim of this study was to compare growth performance, quality of meat, and fatty acid composition in subcutaneous and abdominal fat from ducks fed a diet based on yellow lupin and rapeseed meal, sources of protein alternative to soybean meal (SBM). Ducks were allocated to different dietary treatment groups and reared for 8 weeks (N = 102 per group). Group A received a diet based on SBM, while group B was fed a diet based on yellow lupin with the addition of rapeseed meal. Both groups were divided into two subgroups, of male and female birds. Growth performance parameters and zoometric traits of ducks were monitored during the growth period. After 8 weeks selected birds were slaughtered and dissected (N = 10 per group). Carcass composition was calculated and selected traits of meat quality important for further processing were analysed. Subcutaneous and abdominal fat were collected to analyse fatty acid composition. The alternative diet had no negative effect on ducks’ growth performance parameters and dressing percentage. The replacement of SBM with yellow lupin and rapeseed meal increased n-3 fatty acid content, which is important for consumers. In conclusion, SBM can be replaced with feed containing 60.10% of yellow lupin and 14.00% of rapeseed meal in concentrate. These sources of protein are mainly recommended for small poultry farms, which do not always have access to SBM and prepare poultry feed from their own crops.

. Growth performance parameters in ducks. A ducks fed a diet based on soybean meal; B ducks fed a diet based on yellow lupin and rapeseed meal; ♂♂Males, ♀♀Females. Feed intake (g/duck) Week of production A B Figure 1. Mean daily feed intake (g/duck) during growth. Table 3. Body weight gain (g) in subsequent weeks of duck growth ( x ) (N = 102 ducks per group). A ducks fed an SBM-based diet; B ducks fed a diet based on yellow lupin and rapeseed meal. a,b Means in rows marked with different letters differ significantly (P ≤ 0.05). x Significant group: sex interaction (P ≤ 0.05).  Table 4. Zoometric traits (cm) and body conformation indices (%) of ducks ( x ± SD) (N = 10 ducks per group). A ducks fed an SBM-based diet; B ducks fed a diet based on yellow lupin and rapeseed meal. a,b Means in rows marked with different letters differ significantly (P ≤ 0.05). x Significant group: sex interaction (P ≤ 0.05). www.nature.com/scientificreports/ portion of skin with subcutaneous fat, abdominal fat and the proportion of neck in the carcass; all these parameters except weight of stomach and liver were higher in females (P < 0.05). No significant differences were found between other variables (P > 0.05). There was no significant interaction between variables: treatment group and sex (P > 0.05), except for live body weight of ducks (P = 0.000) ( Table 5). Better cooking loss of breast muscles was found in birds from group B fed a diet based on yellow lupin compared to group A (P = 0.026). Other physicochemical parameters did not differ significantly between groups for both variables (P > 0.05). No interaction effect was found between treatment group and sex (P > 0.05) ( Table 6). Higher content of linolenic acid (C18:3 n-3) and higher content of eicosenoic acid (C20:1 n-9) in subcutaneous fat in group B compared to control ducks fed the SBM-based diet was found. The percentage of n-3 fatty acids (C18:n-3) was significantly higher in group B compared to group A, but the ratio PUFA n-6/n-3 was significantly lower (P = 0.000). Fatty acid composition was also analysed in relation to sex. The content of pentadecanoic (C15:0), palmitoleic (C16:1), heptadecanoic (C17:0), linoleic acid (C18:2 n-6), and polyunsaturated acids (PUFA), including omega-6 and hypocholesterolemic FA (DFA), as well as PUFA/SFA and DFA/OFA (hypercholesterolemic fatty acids) were significantly higher in males than in females (P < 0.05) ( Table 7). No interaction effect between variables was found (P > 0.05). The content of specific fatty acids was also lower in abdominal fat from control group A compared to group B. There were significant differences in the content of linolenic acid (C18:3 n-3), eicosenoic acid (C20:1 n-9) and docosanoic acid (C22:0), and abdominal fat from group A was characterised by a lower content of n-3 acids, but a higher n-6/n-3 ratio (P < 0.05). There were also differences in fatty acid composition, including higher content of heptadecanoic (C17:0), linoleic (C18:2 n-6), docosanoic (C22:0) acids, total PUFA, and the PUFA/SFA ratio in males than in females. The content of palmitoleic acid (C16:1) was higher in females (P = 0.045) ( Table 8). No interaction effect between the diet-sex was found (P > 0.05).

Discussion
Growth performance. In our study the European broiler index (EBI) for ducks ranged from 156.02 (B) to 192.18 (A), and FCR from 3.11 (♂♂ A) to 3.21 (♂♂ B). As Wężyk et al. 16 reported good EBI value is approximately 190. Total feed intake over the 8-week rearing period was approx. 10 kg. Feed intake in control and treatment groups was comparable. Zduńczyk et al. 17 reported that the inclusion of yellow lupin seeds (LL 8, 16 and 24 ) in the diet of turkeys had no negative effect on feed intake. Hejdysz et al. 18 also demonstrated no significant effect www.nature.com/scientificreports/ of narrow-leaved lupin and type of processed seeds on feed intake and FCR in broiler chicken. In our study the mortality rate of ducks was from 1.56% (♂♂ A) to 7.81% (♂♂ B). Presumably, diet was one of the main factors determining the survival of birds. In our study the body weight and weekly body weight gains were lower in ducks from the treatment group compared to control birds (P ≤ 0.05). Kuźniacka et al. 19 reported that the final body weight of ducks fed diets with the inclusion of lupin seeds was comparable to the body weight of ducks fed a balanced SBM-based feed. Biesiada-Drzazga et al. 20 found higher body weight and growth rate in Pekin ducks (♂♀) at 14 and 28 days-old, but lower at 49 days-old compared to the body weight of ducks measured in our study. Mierlita and Popovici 21 reported from their experimental study on Ross 308 broiler chickens the significantly highest body weight in 21-day-old birds fed a standard SBM-based diet and the lowest in the LA 40 group. However, the body weight of 35-day-old broiler chickens was comparable. A 40 and 60% inclusion of lupin flour per feed ration had no adverse effect on the body weight of birds measured at the end of the rearing period (day 42). Daily feed intake in control and treatment groups was comparable. Suchỳ et al. 22 also found no significant differences between the body weight of Ross 308 broiler chickens fed white lupin (LA) seeds LA 6.66-31.03 measured on days 15, 30 and 42 of rearing. In week 2 of life, the highest body weight gain was found in ducks from the control group A (104.39%) and treatment group B (95.37%). The values of body conformation indices, sternum crest length and chest circumference were higher in female ducks. Experiments conducted by Kokoszyński and Bernacki 23 conducted on P44 and P55 ducks revealed lower values compared to those found in our study.
Carcass traits and physicochemical properties of meat. Body weight and carcass weight were significantly higher in control ducks than in ducks from the treatment group. The analysis of carcass quality revealed significant differences in the weight of breast muscles and wings, as well as the proportion of wings in gutted carcass with neck. Kuźniacka et al. 19 reported a slightly higher dressing percentage in ducks fed a balanced feed based on narrow lupin seeds compared to ducks fed mixtures with the inclusion of soybean meal. An experimental study by Witak et al. 24 revealed that the inclusion of yellow lupin seeds in the diet of A44 ducks (2.5, 5 and 10% in weeks 1 to 3 of life, and 7.5, 10 and 15% from week 4 of rearing to slaughter) had no effect on the weight of gutted carcass with neck (g) and the weight and proportion of skin with subcutaneous fat, abdominal fat, breast muscles and leg muscles, or carcass remains. Witak et al. 24 reported a negative (P ≤ 0.05) effect of a 10% inclusion of lupin seeds on the dressing percentage of ducks compared to the control group and other treatment groups. There were also significant sex-related differences in the dressing percentage and the proportion of subcutaneous fat in carcass, which were higher in female ducks, similar to our study. Only the weight of carcass remains was higher in male ducks. The proportion of individual elements of carcass reported by Witak et al. 24 was slightly higher than in our study. In the study by Suchỳ et al. 22 carcass weight and the weight and proportion of breast muscles in carcass was significantly lower in Ross 308 broiler chickens fed a diet with the inclusion of white lupin seeds (LA) compared to the control group (SBM), but the dressing percentage of birds was comparable (70.99% in the control group vs. 73.69% in treatment groups). Diet also had no negative effect on the weight and proportion of neck and legs in carcass. Comparable results were found in our study. Our study found no significant effect of the source of protein in the diet on the weight of carcass elements, but the weight of stomach and liver was significantly higher in males compared to females. In our study the proportion of wings in the total weight of carcass was between 11.89% for female ducks and 12.32% for male ducks. Similar data on the weight of duck wings were reported by Kokoszyński et al. 25 . www.nature.com/scientificreports/ In our study, dietary treatments had no significant effect on the colour of duck breast and leg muscles. Our findings were different from those reported by other researchers [26][27][28] . Other studies 29 found that muscles from turkeys fed LL 16 and LL 24 diets were characterised by a significantly greater yellowness (b*) compared to control birds. An experimental study by Witak et al. 24 found no significant differences in the colour of breast and leg muscles between the control ducks receiving a feed mixture with the addition of SBM, and treatment groups (LL 2.5-15 ) or between males and females, which is consistent with our results.
Moreover, Witak et al. 24 reported that the source of plant protein did not influence the water-holding capacity of breast and leg muscles from broiler ducks. The sex of birds was the only variable with a significant impact on the water-holding capacity of breast muscles and was higher in female than in male birds. On the other hand, Kuźniacka et al. 19 reported higher water-holding capacity of breast and leg muscles from ducks fed a diet with the inclusion of lupin seeds compared to birds receiving a feed mixture with SBM. Wu et al. 30 reported that in Cherry Valley ducks cooking loss was 38.10%, and drip loss was 7.48%. In our study we found no relationship between the sex of ducks or their diet and the pH of breast muscles, which according to literature data is in the range of 5.65-6.20 and depends on the level of glycogen reserves in muscles 27,31,32 . The value of pH decreases post-mortem due to the increased concentration of lactic acid in muscles, and this process determines meat quality. The rapid drop in pH may result in meat paleness, reduced water-holding capacity and an excessively soft structure of meat 33 . In our experiments the post-mortem drop in the pH of breast muscles was normal, which indicates normal metabolic changes in muscles 28 . Krawczyk et al. 29 did not observe any significant differences in the pH of turkey meat measured 24 h post-mortem between the control group (SBM) and treatment groups (LL [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] ). The pH 24 of breast muscles from ducks measured in our study was higher than 5.8. Witak et al. 24 reported slightly lower pH values for breast muscles (5.74 to 5.78) from ducks on an LL diet compared to our findings. Wu et al. 30 reported that the pH of muscles from Cherry Valley ducks was in the range of 5.97 (pH 45 min ) to 5.77 (pH 24h ). The water-holding capacity of duck muscles was comparable between groups and sexes of birds.  www.nature.com/scientificreports/ Lipid content and fatty acid composition of tissues. Fat from lupin seeds is characterised by a high content of unsaturated fatty acids, especially linoleic acid (C18:2 n-6) 34 . Mieczkowska and Smulikowska 35 reported that the dietary inclusion of lupin seeds influenced the level of oleic acid (C18:1 n-9) and α-linolenic acid (C18:3 n-3) in fat from broilers. Diet has a modifying effect on fatty acid composition in somatic lipids, and this improves the nutritional properties of poultry meat. Results of the cited studies correspond with our findings, since we measured significantly higher content of α-linolenic acid in subcutaneous and abdominal fat from ducks on a diet with the inclusion of lupin seeds. Biesiada-Drzazga 31 indicated that the partial replacement of soybean meal with lupin seed meal and rapeseed meal in the diet of White Kołuda geese reared for 10 weeks had a positive effect and was associated with a 2.2% reduction in the total content of SFA, and a 2.85% increase in the total content of MUFA. On the other hand, a significantly lower content of SFA and higher content of UFA was found in skin with subcutaneous fat (by 1.7% on average) and in leg muscles (by 1.2 and 1.7%, respectively) from geese receiving sunflower meal or lupin meal with sunflower meal 32 . Another study by Biesiada-Drzazga 36 revealed no differences in fatty acid composition for breast muscle, leg muscles and skin with subcutaneous fat between geese fed soybean meal and geese fed a diet with an admixture of rapeseed meal. Skin with subcutaneous fat was characterised by a more beneficial fatty acid composition compared to abdominal fat (higher UFA/ SFA ratio), which corresponds with findings by Karpińska and Batura 37 . In oat geese, a dietary inclusion of sunflower seeds (5% in weeks 0 to 3, 9% in weeks 4 to 8 and 14% in weeks 9 to 10) had a positive effect on the content of oleic acid (C18:1 n-9) in abdominal fat, subcutaneous fat and lipids from breast muscles, and on the content of palmitoleic acid (C16:1) in subcutaneous fat and breast muscles 38 . Feeding turkeys with a mixture based on yellow lupin seeds (LL [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] ) influenced fatty acid composition in breast muscles by reducing the level of saturated fatty acids (SFA), including palmitic (C16:0) and myristic acid (C14:0), and increasing the levels of polyunsaturated fatty acids (PUFA): linoleic (C18:2 n-6) and linolenic acid (C18:3 n-3) 29 . Experimental studies by Kiczorowska et al. 39 revealed significantly higher levels of palmitoleic (C16:1) and arachidonic acid (C20:4 n-9), and a Table 8. Fatty acid composition (% of total fatty acids) in abdominal fat from ducks ( x ± SD) (N = 10 ducks per group). A ducks fed an SBM-based diet; B ducks fed a diet based on yellow lupin and rapeseed meal. DFA hypocholesterolemic fatty acids, OFA hypercholesterolemic fatty acids, UFA unsaturated fatty acids, SFA saturated fatty acids, MUFA monounsaturated fatty acids, PUFA polyunsaturated fatty acids; n-3 OMEGA 3, n-6 OMEGA 6, n-9 OMEGA 9. a b Means in rows marked with different letters differ significantly (P ≤ 0.05). www.nature.com/scientificreports/ significantly lower level of myristic acid (C14:0) in abdominal fat from broiler chickens fed with a 50% inclusion of raw peas. This diet was also associated with increased content of palmitic (C16:0) and γ-linolenic acid (C18:3 n-6) in lipids from leg muscles compared to controls fed SBM.
In conclusion, yellow lupin in combination with rapeseed meal can be a source of protein in feed for broiler ducks, as shown by growth performance results and the dressing percentage of the carcass. Growth performance parameters and dressing percentage achieved for ducks fed this diet were comparable to those in birds on the balanced SBM-based diet. The use of balanced feeds with the inclusion of yellow lupin and rapeseed meal has a positive effect on fatty acid composition in abdominal and subcutaneous fat from broiler ducks. Meat from ducks fed this type of diet was characterized by a highest level of hypocholesterolemic and PUFA n-3. Feed with the inclusion of lupin seeds and rapeseed meal can be used in the diet of ducks at small poultry farms operating semi-intensive production, as indicated by the good quality of obtained meat. It is important for small farms, because not everybody has a possibility for own crops of soybean meal. In aspect of consumer market, nowadays is niche of non-GMO products, so it could give a wider choice for potential consumers.

Materials and methods
The research were done with recommendations of directive no. 2010/63/EU and resolution 13/2016 of the National Ethics Committee for Animal Experiments of June 17, 2016. The approval of Ethic Committee was not required. The slaughter of birds was carried out in accordance with the applicable rules on the handling of animals at the time of slaughter, including humane treatment. Also the methods used in the meat quality tests were carried out in accordance with the current and commonly used methods.
Animals and diets. The study was conducted on 204 Cherry Valley English Pekin ducks. Each duckling was weighted on Radwag PS 750/X scales, sexed, and marked with a jiffy wing band. Ducklings were reared in closed facilities under controlled environmental conditions (in accordance with Polish law, ducks were kept to a maximum of 10.5 kg/m 2 ). Air temperature was 27 to 31 °C in week 1, 23 to 29 °C in week 2, and 23 to 26 °C in week 3. The mean relative air humidity was 65%. From the third week of life birds had access to outdoor pens. Birds were divided into two groups (N = 102), each group was in five repetition, and subgroups depending on sex, with 51 males and 51 females in each subgroup. Control birds (A) received a mixture with concentrate containing soybean meal (SBM), and the treatment group (B) received a balanced feed with lupin seeds. The diet of ducks (A, B) was a mixture of 40% of concentrate and 60% of wheat from week 1 to 3 of life, and then 30% of concentrate and 70% of wheat from week 4 to 8. The components of the concentrate and the nutritional value of feed mixture are presented in Table 9. The chemical composition of yellow lupin seeds, cv. Mister, is presented in Table 9. Components of concentrates for ducks. *Vitamin-mineral premix provided per kg diet: Mn, 55 mg; Zn, 50 mg; Fe, 80 mg; Cu, 5 mg; Se, 0.1 mg; I, 0.36 mg; Na, 1.6 g, retinol, 2.48 mg; cholecalciferol 25 μg; DL-ɑtocopherol, 60 mg; cyanocobalamin, 0.012 mg; menadione sodium bisulphite, 1.1 mg; niacin, 53 mg; choline chloride, 1,020 mg; folic acid, 0.75 mg; biotin, 0.25 mg; riboflavin, 5.5 mg. www.nature.com/scientificreports/ Table 10. The chemical composition of rapeseed meal was not presented since it was not the main objective of the study. To supplement nutrients in each treatment group, between weeks 2 and 8 birds received powder product containing amino acids and vitamins (dose 125 g/1,000 L drinking water, Amino-Vitasol WSP, Medivet), as well as SELVITA Vitamin E + Selenium (INVESA) at a dose of 1,000 ml/ 4,000 l drinking water between weeks 5 and 8. Supplementation was the same in all groups. www.nature.com/scientificreports/ Growth performance. Ducks were reared to 8 weeks of age, consistently with the generally used technology for rearing broiler ducks. Birds were reared in a semi-intensive system and received feed ad libitum. Total feed intake was recorded each week in each group, and the mean feed intake was calculated. The body weight of birds in all groups was measured each week (WLC 12/F1/R scales, Radwag) with accuracy to the nearest ± 0.2 g. Weekly feed intake (g) and body weight (g) were used to calculate the feed conversion ratio per kg of body weight gain (FCR, %). The mortality rate was calculated at the end of the duck rearing period. Gathered data were used to calculate the European Broiler Index (EBI) from the formula: where: LBW-live body weight at the end of duck rearing (kg), SR-survival rate (%), AB-age of birds (days), FCR-feed intake per kg of body weight (kg).
Carcass and meat traits. Birds were slaughtered at the age of 8 weeks. Ducks were stunned by electricity, and then their heads were cut off (quick bleeding), according to the requirements. The fasting of birds before slaughter lasted 12 h. Five males and five females were selected from each group, with a body weight close to the mean weight of the same sex individuals in their group. Zoometric traits were measured post-mortem with accuracy to the nearest 1 mm: the length of the abdomen with neck (between the first cervical vertebra and the posterior margin of the ischium), the length of the abdomen (between the shoulder joint and the posterior margin of the ischium), the length of the sternum crest (from the anterior to the posterior margin), the length of the shank (between the ankle joint and the lower posterior surface of the first toe at its base), and the chest circumference (behind the wings, along the anterior margin of the sternum crest and the middle thoracic vertebra). Gathered data were used to calculate body conformation indices (%): compactness index (ratio of chest circumference to abdomen length in cm), massiveness index (body weight in kg to abdomen length in cm), and long-leggedness index (shank length to abdomen length in cm). The carcasses were gutted. Using a pH-meter (CP-401, ELMETRON, with an OSH 2,105 knife electrode) with accuracy to the nearest ± 0.01 pH of breast muscle were done. The electrode was inserted at a 45˚ angle into the right superficial breast muscle. Measurement of carcass pH was after 24 h of cold storage at + 4 °C (pH24). Duck carcasses were dissected at a laboratory using the simplified method described by Ziołecki and Doruchowski 40 . The following elements of carcass were weighed (g): carcass with neck, neck, wings, skin with subcutaneous fat, abdominal fat, breast muscles, leg muscles, carcass remains and offal. Carcass elements were weighed on a WLC 12/F1/R scales (Radwag) with accuracy to the nearest ± 0.2 g, and their proportions in the weight of gutted carcass with neck were calculated in %. The weight of gutted carcass without offal (g) and body weight before slaughter (g) were used for the calculation of the dressing percentage (g) of birds according to the formula: For physicochemical analysis 10 (5 males and 5 females) breast and legs muscles from each group were taken. Each piece was assigned by number. Breast and leg muscles were analysed without skin. The outer right part of breast muscles, leg muscles and skin were analysed for colour using a colorimeter (CR 400, Minolta) in the CIE L* a* b* system 41 . Drip loss from breast and leg muscles was measured. For that purpose, right breast muscles and leg muscles were sampled and weighed on a PS 750/X scales (Radwag) with accuracy to the nearest ± 0.1 g, attached to a special stand and left for 24 h in a cold room at 4 °C, weighed again, and drip loss in % was calculated based on the difference in muscle weight before and after storage 42 . The water-holding capacity of breast and leg muscles was analysed with the Grau and Hamm method 43 . Muscles were disintegrated (K35 mincer, Electrolux), and 280 to 320 mg samples were weighed with accuracy to the nearest ± 0.01 mg (Radwag PS 750/X scales), wrapped in filter paper, placed between two glass plates and kept under 2 kg pressure for 5 min. Samples were weighted again after 5 min. Water-holding capacity (%) was calculated as the ratio of sample weight after pressing to its weight before pressing (mg). Cooking loss from breast and leg muscles (without skin) was analysed. Muscles were disintegrated in a mincer (K35, Electrolux) and 20 g samples were weighed (Radwag PS 750/X). Samples were wrapped in sterile gauze, tied with string, placed in an 80 °C water bath (ADVERTI) for 30 min, and weighed again (PS 750/X). The difference in weight was used to calculate heat-induced leakage from muscles (%). Carcass and meat quality analysis were done according to the methods described by Biesek et al. 9 .
Fatty acid composition. Samples of subcutaneous and abdominal fat from ducks were preserved, frozen (− 18 °C), lyophilized (Alpha plus apparatus, Donserv) and analysed for fatty acid composition.
Fat was extracted using a technique proposed by Folch et al. 44 , with a mixture of chloroform and methanol (2:1 v/v) on a laboratory shaker. The samples were filtered and left for 24 h for decantation. Fatty acid methyl esters were prepared according to the PN-EN ISO 12,966-2 standard 45 . The fat was dissolved in isooctane and transmethylated with a solution of potassium hydroxide in methanol. Then neutralization of potassium hydroxide with sodium sulphate was carried out. The esters were salted with sodium chloride solution.
Saponified fatty acid esters were separated using a gas chromatograph (type 7,890 B, Agilent Technologies) with an MSD 5977A detector and an autosampler. A capillary column DB-225 MS 60 m × 0.25 mm × 0.25 µm was used for analysis. Analytical parameters were as follows: injection port temperature (split mode 1:100) 230 °C; transfer line temperature 230 °C, ion source temperature 230 °C; quadrupole temperature 150 °C, mode: SIM, ionization type: EI. Oven temperature settings were: 70 °C-increase 0.0 °C/min-hold time 0.0 min; 210 °Cincrease 7.0 °C/min-hold time-65.0 min. Carrier gas-helium (flow rate 1.0 ml/min; volume of injected sample 1.0 µl). Fatty acid methyl esters were identified using the Supelco 37 standard FAME Mix component. www.nature.com/scientificreports/ The individual fatty acids were calculated as the percentage of the total fatty acids identified. Method of fatty acid composition analysis was done similarly, as Kuźniacka et al. 6 described.
Statistical analysis. Data were processed using Statistica 12.5 PL software, 2017. Mean values for all analysed parameters and their standard deviations (SD) and coefficients of variation (v) were calculated. A two-way model of ANOVA was used to analyse variability (variable 1-dietary treatment group, variable 2-sex). The significance of differences was verified using the Tukey test. Interactions between experimental variables were determined. The significance of differences was adopted at P ≤ 0.05. Production results were calculated for the whole flock of ducks, but the quality of meat traits were calculated for chosen ducks for the slaughter.
Ethics. The research were done with recommendations of Directive No. 2010/63/EU and resolution 13/2016 of the National Ethics Committee for Animal Experiments of June 17, 2016. The approval of Ethic Committee was not required. The slaughter of birds was carried out in accordance with the applicable rules on the handling of animals at the time of slaughter, including humane treatment. Also the methods used in the meat quality tests were carried out in accordance with the current and commonly used methods described in the Material and methods section.