Performance of Slow-Growing Male Muscovy Ducks Exposed to Different Dietary Levels of Quebracho Tannin

Simple Summary Different inclusion levels of Quebracho tannin (QT) in the diet of growing male Muscovy ducks of a slow-growing type were explored under free-range conditions. As a result of the dietary treatments tested in this trial, the growth performance or the total blood proteins were not affected. By contrast, dietary QT did not lead birds to produce less moist excreta, as observed in other bird species, as a desirable aspect for intensively raised poultry. A marked improvement in carcass yield was observed as a desirable economic trait in the extensive slow-type duck farming system. Abstract The study of the nutritional effects of tannins is complex due to the large chemical diversity; consequently, in poultry nutrition the biological responses may vary greatly. The aim of the present study was to evaluate the effect of different levels of dietary Quebracho tannins (QT) on growth and production performance in slow-growing type Muscovy ducks. For this purpose, a 42-d trial was carried out on 126 male ducks (42-d old at start), fed on three levels of dietary QT inclusion in the diet (0% as control diet, vs. 1.5% vs. 2.5% on an as fed basis). Birds were reared under free-range conditions. A linear increase in feed intake as a function of QT inclusion in the diet was observed (p < 0.05). No difference as to final body weight, overall average daily weight gain (ADG) and total feed conversion ratio (FCR) in relation to dietary treatments was observed. Carcass yields were positively improved in QT birds (p < 0.05). No adverse responses were recorded in total blood protein and liver weight. Dietary QT might be safely used up by to 2.5% in 42- to 84-d aged male Muscovy ducks.


Introduction
Tannins are plant secondary metabolites and represent the fourth most abundant group of secondary compounds of the plant kingdom, after cellulose, hemicellulose, and lignin [1]. Tannins are polyphenolic compounds, located mainly in vacuoles of the vegetal cell or in waxes, where they do not interfere with plant metabolism. They are found in many parts of the plant, such as fruits, leaves, bark, and wood [2], and in common foodstuffs, for example grapes, strawberries, blackberries,

Birds and Diets
The experimental protocol (prot. no. 814715) was approved by the Bioethical Committee of the University of Turin (Italy). A feeding trial was carried out using one-hundred and twenty-six 42-d old males of Muscovy duck (Cairina moschata domestica, Linnaeus 1758), from a local slow-growing type, preserved since over 40 years at the Avian Conservation Center, which belongs to the Department of Veterinary Science of the University of Pisa (Italy). Birds were individually identified by a wing tag, weighted, and housed randomly in 18 roofed outdoor pens (6x3m) with sandy floor. Each pen housed seven animals.
Birds were allotted to the following experimental dietary groups: the control group (QT0) received the basal diet (Table 1); QT dietary groups received the same basal diet of QT0 with the inclusion of increasing QT levels at an amount of 1.5% and 2.5% on an as-fed basis, respectively. The QT is a commercial animal feed additive, extracted from the heartwood of Schinopsis sp., (MGM-S ® , Unitan SAICA, Buenos Aires, Argentina). This product is available as a fine powder with 58% tannins, 20% phlobaphenes, 14% non-tannic compounds and 8% water and a polymerization degree of 6-7. Due to previous experiences in pheasants [20] and laying hens [21], the maximum safe level of QT to be used in diets for 42-to 84-d old Muscovy ducks is 2.5% on an as-fed basis, therefore, experimental groups were defined as group QT1.5 (2.6% MGM-S powder on top, corresponding to 1.5% QT) and QT2.5 (4.3% MGM-S powder on top, corresponding to 2.5% QT). For each period, the basal diet met or exceeded NRC [33] duck's nutritional requirements and was formulated according to INRA [34] feed ingredient nutritional composition. Six replicates were assigned to each dietary treatment. Water and feed were supplied ad libitum. The chemical composition of the basal feed was analyzed in duplicate [35] and mean results are displayed in tables.

Growth Performance and Feed Conversion Efficiency
Growth performance was evaluated biweekly during the age interval 42-84 d. Individual live body weight (LBW) and feed intake per pen were recorded to calculate the average daily weight gain (ADWG), the average daily feed intake (ADFI) and the feed conversion ratio (FCR).
The LBW in each group were fitted by using LAB Fit-Curve Fitting Software [36]. To trace the first part of the growth curve, individual LBW, previously recorded, of all birds at 0, 14 and 28-d age was considered. The same data were then used in each of the three growth curves (QT0, QT1.5 and QT2.5). The Gompertz function resulted the best fitting growth curve. The following equation describes the curve where W (t) is the weight in grams of duck at time t, W ¥ is the asymptotic final LBW, k is the constant expressing the rate of approach to W ¥ , and t • is the age at the inflexion point where maximum rate of growth is achieved. The maximum rate of growth for each group was calculated by the differential equation of 1 • grade derived from Gompertz The values of W ¥ , k and t • are the same described above for the growth curve, W'(t) is the daily weight increase in grams at time t [37].

Blood Sampling
Blood samples from 24 birds (84 d old) from each dietary group (four birds/pen) were collected during slaughter procedures into K 3 EDTA 2mL vacuum tubes. Total blood protein (TBP) was determined according to Salamano et al. [38].

Excreta Sampling and Dry Matter
Excreta sample collection was carried out on 24 ducks (81 d old) per dietary treatment, following the procedure described by Gariglio et al. [39]. Briefly, the birds were kept and temporary housed (1 h) in cages (two birds in the same pen/cage). Excreta samples free from debris and foreign substances were collected from a tray placed under each cage. A total of 12 pooled samples per dietary treatment were collected at each sampling. The dry matter (DM) of excreta was determined according to AOAC [35].

Slaughter Traits and Organs Evaluation
Twelve birds (84 d old) per dietary treatment were randomly chosen and fasted for 12-h, to be finally weighed and slaughtered in a commercial abattoir. Birds were electrically stunned prior to bleeding. All bird carcasses were mechanically plucked after immersion in hot water. Evisceration was performed manually. Fresh absolute weight of liver and gizzard were determined, and relative weight expressed as percentage of live body weight (LBW). Head-neck and feet were removed and weighed separately; the respective relative weights were calculated as percentage of LBW. Hot weight of ready-to-cook carcass (RCC) was recorded and expressed as percentage of LBW. The small intestine and caeca were unfolded on a flat surface to obtain a fully extended viscera and allow linear length measurement; the respective lengths were expressed as cm per 100g of LBW. Carcasses were stored in a cool chamber at 4 ± 0.5 • C for 24 h, then chilled RCC weights were determined and expressed as percentage of LBW. Finally, m. pectoralis and legs + thigs were separated from the RCC and weights were determined and expressed as percentage of LBW.

Statistical Analysis
The statistical analyses were performed using the SPSS software [40]. Shapiro-Wilk's test established normality or non-normality of distribution, and the assumption of equal variances was assessed by means of Levene's homogeneity of variance test. The experimental unit was the pen. The collected data were tested by means of one-way ANOVA, using polynomial contrasts to test the linear and quadratic responses to increases in the QT inclusion levels in the diet. For all tests, p < 0.05 was considered statistically significant. Results were expressed as the mean and standard error of the means (SEM). A statistical trend was considered for p < 0.10.

Growth Performance and Feed Conversion Ratio
Throughout the experimental period, no mortality was experienced. Bird performance is summarized in Table 2. Overall LBW was not influenced by dietary treatments (p > 0.05), except for 70-d age, where a quadratic response was observed, and the minimum shown by QT1.5 group (p < 0.05). The ADFI resulted directly proportional to the concentration of QT during the 56-70 period, showing a linear (p < 0.01) and a quadratic (p < 0.05) response. A linear increase (p < 0.05) was also observed when considering the whole experimental period (42-84 d). The FCR was not affected by dietary QT, considering both the single age periods, and the whole experimental period.
According to the calculated Gompertz parameters (Table 3), birds reached the inflection point when this trial began (42-d age). Curve parameters were close to each other (Table 3).

Total Blood Proteins
TBP in 84-d old birds are reported in Table 4. The dietary QT inclusion did not influence the level of TBP (p > 0.05). In birds, normal TBP concentrations ranged between 3.5-5.5 g/dl [41]. Results in this study placed all birds within this range.

Excreta Dry Matter Evaluation
Excreta dry matter content in 81-d old birds is reported in Table 5. The dietary QT inclusion did not influence this parameter (p > 0.05).

Slaughter Traits
Slaughter traits in 84-d old male Muscovy ducks are reported in Table 6. The dietary QT inclusion had no influence on the development of breast muscles and Legs + thigs (p > 0.05). Both RCC, hot and chilled, were positively influenced by dietary treatment, showing a linear (p < 0.05) and a quadratic response (p < 0.01). Dietary treatment affected linearly the Head-Neck mass, which decrease by the increase level of dietary QT (p < 0.05). All other slaughter traits were not affected by the dietary treatment (p < 0.05), except for the SIL, which evidenced a quadratic response, with the minimum value expressed by the lowest QT inclusion level (p < 0.01).

Discussion
The increasing interest in the use of some plant-derived molecules with biological properties offers the chance to exploit several beneficial effects due to their numerous properties [9,42]. Against this background, more information seems to be needed about their safe use in different species, to achieve the desired effects. From this perspective, and due to previous experiences in pheasant [20] and pullets [21], the maximum safe QT level to be used in diets for 42 to 84-d old Muscovy ducks was set to 2.5%, on an as-fed basis. At this QT amount, no adverse effects were observed in birds of this trial, as supported by the TBP levels and liver weight. On the contrary, in a previous study involving slow-growing pullets [21], adverse effects were observed in 35-d old birds fed 3% dietary tannins supplementation, resulting in 2% of QT as the safe inclusion level. It is widely known that tannins bind and precipitate proteins, which compromises absorption at intestinal level. This situation might lead to protein deficiency in the bloodstream [9]. Ducks in this study fed on QT diets did not exhibit a decrease in TBP circulating value. Additionally, our results agree with those obtained by Gariglio et al. [43] in 50-d old female Muscovy ducks.
As to liver, weight variations can be useful signs of the presence of hepatic injuries [44]. In this trial, liver weights in QT birds were similar across birds from all experimental groups, at each QT inclusion rate tested. Furthermore, at the end of the experimental period, birds of QT1.5 and QT2.5 groups reached a similar LBW at slaughter. As demonstrated also by the growth curve trends, which were almost identical, it can be stated that the tested amounts of QT included in the diet did not negatively impact the production performance of slow-growing male Muscovy ducks. In agreement with other authors, who reported duck's growth curves using Gompertz model [45,46], we also found this model as the best fitting in our trials (R 2 = 0.98-0.99).
According to the QT inclusion levels tested, a different reaction in birds was noted regarding feed intake. Birds of the QT1.5 group needed a longer time to get used to the change in diet, as demonstrated during the first days after dietary QT inclusion with an evident decrease in the ADWG and ADFI. Birds of the QT2.5 group, instead, seemed to cope with the QT by increasing the ADFI, which was higher during the whole experimental period. Such differences in response to different levels of dietary QT inclusion might be explained by the natural feeding behavior of the Muscovy duck, for which a large part of foraging activity takes place in fresh water [32]. In captivity, when ducks find a hard, unusual or an unknown tasty food, they use to soak it into water in order to render it "eatable". It could be argued that such behavior of soaking feeds was performed with the purpose of expressing the innate feeding behavior. With the increasing QT, the feed taste becomes reasonably astringent and birds tended to put the feed into the water with a consistent higher consumption. In QT1.5 group, instead, astringency perception presumably following a lower QT inclusion amount, might have not led birds to adopt coping solutions, for which a lower feed intake was also observed. Elkin et al. [47] reported no growth depression in ducks fed sorghum rich in tannins. Same authors also suggested that the innate feeding behavior of ducks could be a key factor in the interpretation of such a response.
In this trial, the FCR calculated throughout the experimental period and the final LBW were not affected by dietary QT at both tested levels. However, in QT1.5 birds, an initial retarded growth occurred, with a successive compensatory growth, evident especially during the last 15 days (70-84). This compensatory growth trend was also reported by other authors in broilers reared under feeding restriction, followed by ad libitum feed provision [48]. A decrease in growth rate, lower LBW, lower ADFI and higher FCR were reported in male Muscovy ducks fed low-or high-tannin-sorghum diets [49]. These authors also reported no differences in these parameters in Muscovy males fed on diets 50/50 high/low tannin-sorghum enriched with L-Methionine, attributing the reduction in the negative effect of tannin on protein availability to this amino acid.
Unexpectedly, the use of dietary QT did not lead birds to excrete dryer droppings as observed in other bird species, namely chickens [21] and pheasants [20], but rather the opposite effect could be observed. This result might be attributed to the high daily water intake in ducks, which did not allow the QT to have a "drying" effect. As some authors reported, 800 ml/d was the water intake in 14 to 42-d old Pekin ducks [50], while in 20-week old White Leghorn hens the intake was 228 ml/d [51], thus, threefold lower than the intake in ducks.
The dietary QT did not negatively affect carcass traits in 84-d old male ducks, rather, it led to improved yields in both the hot and chilled RCC of QT birds, which was heavier than QT0. Higher RCC yields were also reported in male Muscovy ducks of the same strain [52], fed on diets with 50/50 high/low tannin-sorghum [48]. In slow-growing chickens, instead, the RCC yield pointed out no differences [21], as well as in broiler chickens [26] and in female pheasants [20].
Wild birds feeding on diets rich in tannins developed larger intestines and caeca, and heavier gizzards [53], however, this could be also induced by a different physical form of natural feeding sources, rich also in more fibrous nutrient content [54]. In this trial, small intestine length was influenced by the diet. Nevertheless, this influence resulted in the opposite to what we would have expected, with QT1.5 birds having a shorter small intestine. In contrast, no effect was observed in growing female pheasants after a 60-d QT diet [20], and longer small intestine length after a 4-week trial in adult grey partridges was observed by Liukkonen-Anttila et al. [53]. Further research is needed to clarify this result.

Conclusions
In view of the parameters tested in this feeding trial, no adverse effects on health nor on the production performance of birds could be observed. The increasing amount of QT in the diet was found to be linked to the higher intake of feed by the ducks of the different dietary groups, likely due to the capability of dietary quebracho tannins to stimulate the innate feeding behavior. Excreta quality was not influenced by the amount of tannins in the diet, probably due to a conserved water consumption via soaked feed. Finally, the slaughtering performance was satisfactory and improved by QT supplementation, thus the inclusion up to 2.5% of Quebracho tannins in the diet can be safely used in slow-growing type male Muscovy ducks.