The effect of grouping one-day-old chicks by body weight on the uniformity of broilers

2013 Poultry Science Association, Inc. The effect of grouping one-day-old chicks by body weight on the uniformity of broilers Roberto Montanhini Neto, 1 Diego Surek, Chayane da Rocha, Fabiano Dahlke, and Alex Maiorka Veterinary Sciences, Federal University of Paraná, Curitiba, Brazil 80035-050 Primary Audience: Flock Supervisors, Field Operations Managers, Researchers DESCRIPTION OF PROBLEM The average weight and homogeneity in weight of a flock of broilers are important parameters in poultry production [1, 2]. Uniform flocks with adequate weights present several management advantages: (1) poultry are handled in large groups and can be simultaneously exposed to changes in management (lighting, feeding, and environment), (2) more effective overall performance is obtained, and (3) the birds achieve a level of performance that approaches their maximum genetic potential [3]. The emphasis on broiler uniformity has frequently been dictated by the last step in the production line: the processing facility. Automation SUMMARY In this study, we evaluated the effect of the grouping of 1-d-old chicks according to initial BW on the subsequent homogeneity and distribution of the weight of the broilers at harvest. Two treatments (placement methods) were tested: in one treatment (random), the chicks placement was at random and not grouped by initial weight; in the other treatment (homogeneous groups), the chicks were grouped according to their initial weight. The broilers were individually weighed upon placement and again at 21 and 42 d of age. The lowest dispersion of the weight at placement was observed for the homogeneous groups, but the dispersion did not differ between the treatments at the subsequent weighings. Despite the difference in placement, the distributions of the weight data for the 2 groups did not differ between 21 or 42 d of age. Based on these results, the grouping of chicks by weight does not produce more uniform broilers at the end of the production period. Key words: distribution, dispersion, homogeneity, initial weight 2013 J. Appl. Poult. Res. 22 :245 250 http://dx.doi.org/10.3382/japr.2012-00636 used for processing requires a uniform carcass size to meet the standards for the quality of the poultry received by the processing unit [4]. Often, the quality of the management procedures used by the producer affects the uniformity obtained in a flock. Indeed, the homogeneity of harvest weight may be strongly influenced by events occurring during the production period [5]. Because increases in uniformity during the production period are unlikely, several authors view the uniformity of chicks at placement as a key prerequisite for obtaining a uniform final product [6]. At placement, the weight of the chicks is an economically important feature because every 1-g increase in the initial weight 1 Corresponding author: roberto.neto@ufpr.br

246 JAPR: Research Report results in a 7 to 13 g increase in the harvest weight of the broilers [7]. In contrast, several researchers have shown that improvements in the uniformity of chicks or segregation by initial weight, especially in chicks from small eggs or young flocks, may reduce the degree of competition, thus decreasing the level of mortality and increasing the homogeneity of the weight of the broilers entering the processing unit [8 11]. Chick flocks that lack sufficient uniformity cannot be properly managed. This situation results in lower growth, increased feed intake, and higher mortality during the first weeks [12]. Based on studies with several broiler flocks, the CV of individual BW from a single flock of breeder hens may vary between 7.5 and 10.7%, with an average of approximately 9.25%. Accordingly, the 95% CI of the CV would include values from 8.5 to 10% [11, 13, 14]. In addition, based on a survey, broiler flocks with higher values of the CV (reduced uniformity) experienced higher mortality rates. This measure of dispersion can indicate the zootechnical quality of the Table 1. Composition of the experimental diets Item chicks; a lower CV corresponds to higher chick quality [15]. The purpose of the present study was to assess the effect of the grouping of chicks by initial weight on the subsequent homogeneity and distribution of weight of the broilers. MATERIALS AND METHODS The experiment was conducted in the experimental poultry house of the Federal University of Paraná in southern Brazil. The study included 720 one-day-old male broiler chicks of the Ross 508 strain from 47-wk-old hens of the same flock. The chicks were housed in an experimental facility divided into 24 pens, with 30 birds per experimental unit. The management procedure used was that recommended by the Brazilian poultry industry [16], with feed and water ad libitum. Table 1 shows the calculated composition of the diets of the 3 phases. The experimental units were randomly distributed into 2 treatments. In the first treatment, the chicks were placed at random, without any Feeding period d 1 to 21 d 22 to 35 d 36 to 42 Ingredient (g/kg) Corn 553.0 626.9 632.2 Full-fat soybean 194.0 180.0 159.0 Soybean meal 181.0 102.0 148.0 Meat and bone meal 32.0 26.0 22.0 Poultry fat 4.0 20.0 Poultry by-product meal 15.0 19.0 Feather meal 5.0 24.0 Limestone 6.5 4.5 7.0 NaCl 4.0 4.0 4.0 Vitamin and mineral premix 1 4.0 4.0 4.0 Methionine hydroxy-analog 3.2 2.4 2.2 l-lysine HCl 1.6 2.6 1.0 Choline chloride 0.7 0.7 0.7 Calculated nutritional content Apparent ME (kcal/kg) 3,100 3,200 3,250 CP (g/kg) 220.0 200.0 181.8 Digestible lysine (g/kg) 12.9 11.4 10.3 Digestible methionine (g/kg) 6.2 5.1 4.8 Digestible methionine + cystine (g/kg) 9.8 8.8 7.9 Calcium (g/kg) 10.0 8.7 7.7 Available phosphorus (g/kg) 4.5 4.1 3.6 Sodium (g/kg) 2.1 2.2 2.0 1 Contribution per kilogram of feed: vitamin A = 8,000 IU; vitamin D 3 = 2,400 IU; vitamin E = 16.65 mg; vitamin K 3 = 1.5 mg; vitamin B 1 = 0.6 mg; vitamin B 2 = 2.36 mg; vitamin B 6 = 0.6 mg; vitamin B 12 = 1,320 µg; biotin = 0.15 mg; pantothenic acid = 9.32 mg; niacin = 30.12 mg; folic acid = 1.42 mg; selenium = 0.65 mg; iodine = 0.35 mg; iron = 57.72 mg; copper = 12.30 mg; zinc = 141.48 mg; manganese = 173.0 mg.

MONTANHINI Neto et al.: CHICK WEIGHT AND UNIFORMITY 247 Table 2. Comparisons of the average BW (grams) of chicks aged 1, 21, and 42 d for the different placement methods of 1-d-old chicks Placement method 1 d 21 d 42 d Random (n = 360) 43.1 (±3.8) 799.6 (±77.6) 2,381.4 (±223.4) Homogeneous group (n = 360) 44.5 (±0.8) 809.2 (±80.9) 2,421.8 (±230.1) P-value 1 <0.001 0.049 0.032 1 P-values lower than 0.05 indicate significant differences (Mann-Whitney test). grouping by initial weight. In the second treatment, the chicks were separated into homogenous groups according to their initial weight. For grouping, cutoff values were based on tertiles of the initial weight. The chicks were grouped in pens for this treatment according to the following categories: light (lower tertile), medium (intermediate tertile), and heavy (upper tertile). In this treatment, the birds with weights classified as outliers were excluded. An outlier was defined as a BW beyond a range of 3 or more standard deviations above or below the average weight. The broilers were individually weighed at placement and again at 21 and 42 d old. The central tendency, dispersion, and distribution of the data were used for statistical comparisons [17]. The tendencies were compared with a Mann-Whitney test [18] because the data did not meet the assumption of normality according to a Shapiro-Wilk test [19]. The data were expressed as the arithmetic mean of the weights. To compare the percentage difference in mean weight between treatments for the results of each weighing, a Chi-squared test and a Fisher exact test were used [20]. The dispersion was evaluated by comparing the variances with a Levene test [21]. The dispersion data were presented in terms of a measure of relative dispersion, the Pearson CV, which was expressed as a percentage. The values of the relative dispersion for each weighing were evaluated and their association tested with Spearman correlation coefficients. The distributions were compared with a Kolmogorov-Smirnov test and presented as histograms. To determine the position and the flattening of the data distributions, the kurtosis and asymmetry coefficients [22] were calculated and compared with a Wald-Wolfowitz test [23]. The significance level was set at 5% for all statistical tests used in this study; all tests were performed with Statistica software, version 8.0 [24]. RESULTS AND DISCUSSION The central tendencies differed significantly (P < 0.05) between treatments for all weighings (Table 2), with higher weight values observed for the homogeneous groups. However, the percentage difference between the treatments for the 3 weighings was found to be statistically similar (P > 0.05). If the treatments were compared in terms of the weighings on d 21 and 42 with the placement weight as a covariate, the differences were not significant (P > 0.05). No significant differences (P > 0.05) were observed in feed consumption and FCR between the treatments on the 3 phases of the experiment. Therefore, the difference in weights for the weighings on d 21 and 42 is directly related to the difference in weights at placement and not to the treatments. Based on the literature, in general, if chicks are heavier at birth, heavier broilers result at harvest [7]. The broiler weight at harvest is directly proportional to the chick weight at 7 d, resulting in a high positive correlation between the weight at placement and at harvest [25]. Table 3. Comparison between the CV (%) of chick body weight data obtained at placement and at d 1, 21, and 42 for the different placement methods of day-old chicks Placement method d 1 d 21 d 42 P-value Random (n = 12) 8.79 9.73 9.39 0.684 Homogeneous group (n = 12) 1.69 b 10.02 a 9.52 a <0.001 P-value <0.001 0.334 0.631 a,b P-values lower than 0.05 and values with different letters in the same row or column indicate significant differences (Levene test).

248 JAPR: Research Report Figure 1. Comparisons of the BW distributions of data obtained for chicks at d 1, 21, and 42 for the different placement methods of 1-d-old chicks.

MONTANHINI Neto et al.: CHICK WEIGHT AND UNIFORMITY 249 Table 4. Comparison between the kurtosis and asymmetry coefficients of chick BW data obtained at d 1, 21, and 42 for the different placement methods of 1-d-old chicks Placement method d 1 d 21 d 42 P-value Kurtosis coefficient Random (n = 12) 0.934 0.081 0.119 0.717 Homogeneous group (n = 12) 2.223 a 1.065 b 1.012 b 0.002 P-value 0.006 0.139 0.334 Asymmetry coefficient Random (n = 12) 0.128 0.010 0.090 0.125 Homogeneous group (n = 12) 0.340 a 0.156 b 0.123 b 0.038 P-value 0.044 0.759 0.119 a,b P-values lower than 0.05 and values with different letters in the same row or column indicate significant differences (Wald- Wolfowitz test). A lower (P < 0.05) dispersion of weight at placement was observed for the homogeneous groups. Nevertheless, no difference in dispersion was found between the treatments for the subsequent weighings (Table 3). Note that the chicks came from the same flock and from breeder hens of the same age. Furthermore, no significant correlation (P > 0.05) was detected between the CV at placement and the CV for the subsequent weighings. These results are consistent with reports in the literature that describe low correlations between the uniformity of the weights of 1-d-old chicks and the uniformity of broiler weights at 42 to 44 d [6, 7, 26 29]. In these studies, several factors have been cited as determinants of the homogeneity of BW at harvest, but the consensus is that the individual capacity of each bird to gain weight over its life is independent of the initial weight grouping. Despite the initial grouping reflected in the low dispersion of the homogeneous groups on d 1, the CV for the broilers on d 21 or 42 were similar to those observed for the treatment with random placement (Table 3). In the present experiment, the weight homogeneity of the broilers midway through the experiment and at the end of the experiment was not related to the weight homogeneity at the time of placement. These findings demonstrated that the uniformity of the flock does not depend on the initial weight uniformity of the chicks if this initial weight uniformity is achieved by equalizing the weight of the chicks at the time of placement. In practice, other factors related to broiler management, such as the intake of water and feed, the formation of groups and hierarchies, the maintenance of a suitable density, an appropriate placement environment, as well as immunity and health, appear to be more important for maintaining the uniformity of the flock until the end of the production cycle [6, 30]. The distribution of weights differed significantly between treatments (P < 0.05) on the day of placement. One of the goals of the homogeneous groups treatment was to exclude chicks with extreme weights. However, the distribution of weights did not differ between the treatments on d 21 or 42 (Figure 1). The CV did not differ significantly among the weighings (P > 0.05) for the random treatment; however, the CV was lower (P < 0.05) on the day of placement than on the subsequent weighings for the homogeneous groups. Therefore, grouping by initial weight does not ensure that broilers subsequently grow without achieving extreme weights. Based on the analysis of the difference between treatments in the kurtosis and asymmetry coefficients, both measures of the distribution of the data differed (P < 0.05) only on the day of placement. The coefficients were similar (P > 0.05) on the other weighing days (Table 4). Compared among the weighings, the measures were similar for the random treatment; however, the coefficients for the homogeneous groups indicated differences (P < 0.05) between the day of placement and the 2 subsequent weighing days. The coefficients did not differ significantly between d 21 or 42 (P > 0.05), though. The results confirm and strengthen the results based on the CV. The reason for this outcome is that the kurtosis coefficient is a descriptive evaluation of the data distribution, but also serves to assess the dispersion of the data in terms of the level of flattening of the distribution [22].

250 JAPR: Research Report CONCLUSIONS AND APPLICATIONS 1. The grouping of 1-d-old chicks by BW at placement does not ensure a greater uniformity of the broilers at the end of the production period. 2. The difference between mean values of BW at placement directly influenced mean values of the weights at subsequent ages. 3. The exclusion of individuals whose weight at placement was classified as an outlier did not guarantee that the flocks would show a lower dispersion of weights at harvest. REFERENCES AND NOTES 1. Dalanezi, J. A., A. A. Mendes, E. A. Garcia, R. G. Garcia, J. Moreira, and I. C. L. A. Paz. 2005. Effect of broiler breeder age on performance and carcass yield of broiler chickens. Arq. Bras. Med. Vet. Zootec. 57:250 260. 2. Molenaar, R., I. A. M. Reijrink, R. Meijerhof, and H. Van den Brand. 2008. Relationship between hatchling length and weight on later productive performance in broilers. W. Poult. Sci. J. 64:599 604. http://dx.doi.org/10.1017/ S0043933908000226. 3. Kosba, M. A., H. S. Zeweil, M. H. Ahmed, S. M. Shabara, and A. A. Debes. 2010. Selection for uniformity in Alexandria local chickens: 2. Correlated response for productive and reproductive traits. Egypt. Poult. Sci. 30:114 136. 4. Griffin, A. M., R. A. Renema, F. E. Robinson, and M. J. Zuidhof. 2005. The influence of rearing light period and the use of broiler or broiler breeder diets on forty-twoday body weight, fleshing, and flock uniformity in broiler stocks. J. Appl. Poult. Res. 14:204 216. 5. Judice, M. G., J. A. Muniz, and R. Carvalheiro. 1999. Evaluation of coefficient of variation in experimentation with swine. Ciên. Agrotec. 23:170 173. 6. Rocha, J. S. R., L. J. C. Lara, N. C. Baião, S. V. Cançado, M. V. Triginelli, and J. F. C. Leite. 2008. Effect of egg classification prior to setting on broiler chicken uniformity, performance and meat yields. Arq. Bras. Med. Vet. Zootec. 60:1181 1187. 7. Wilson, H. R. 1991. Interrelationships of egg size, chick size, posthatching growth and hatchability. W. Poult. Sci. J. 47:5 20. 8. Joseph, N. S., and E. T. Moran Jr.. 2005. Effect of flock age and post emergent holding in the hatcher on broiler live performance and further-processing yield. J. Appl. Poult. Res. 14:512 520. 9. Gardiner, E. E. 1973. Effects of egg weight on posthatching growth rate of broiler chicks. Can. J. Anim. Sci. 53:665 668. 10. Hearn, P. J. 1986. Making use of small hatching eggs in an integrated broiler company. Br. Poult. Sci. 27:498 504. 11. Wilson, H. R., and M. E. Suarez. 1993. The use of egg weight and chick weight coefficients of variation as quality indicators in hatchery management. J. Appl. Poult. Res. 2:227 231. 12. Van der Ven, I. L. 2005. Maximizing uniformity through top-level hatchery practice. W. Poult. J. 21:1 3. 13. Bondarenko, U. B. 1989. Variation in egg and day old chick weights of nine fowl species. Inst. Poult. Res. Bull. 26:6 10. 14. Shalev, B. A., and H. Pasternak. 1995. Incremental changes in and distribution of chick weight with hen age in four poultry species. Br. Poult. Sci. 36:415 424. 15. Le Turdu, Y., P. Drouin, J. Y. Toux, J. Josse, M. Guittet, J. P. Picault, G. Bennejean, P. Quemeneur, and N. Hamet. 1984. Production du poulet de chair destine a Pexportation. Bull. D inform. Stat. Exp. D avic. Plouf. 24:39 47. 16. Mendes, A. A., I. A. Naas, and M. Macari. 2004. Produção de frangos de corte. Ed. FACTA, Campinas, Brazil. 17. Kaps, M., and W. R. Lamberson. 2009. Biostatistics for animal science. CAB International, Oxfordshire, UK. 18. Harding, E. F. 1984. An efficient, minimal-storage procedure for calculating the Mann-Whitney U, generalized U and similar distributions. Appl. Stat. 33:1 6. 19. Royston, J. P. 1982. An extension of Shapiro and Wilk s W test for normality to large samples. Appl. Stat. 31:115 124. 20. Upton, G. J. G. 1992. Fisher s exact test. Stat. Soc. 155:395 402. 21. Schultz, B. B. 1985. Levene s test for relative variation. Syst. Biol. 34:449 456. 22. Ferreira, D. F. 2005. Estatística básica. Ed. UFLA, Lavras, Brazil. 23. Thas, O. 2010. Comparing distributions. Springer Science Business Media, Ghent, Belgium. 24. Hill, T., and P. Lewicki. 2007. Statistics: Methods and Applications. StatSoft Inc., Tulsa, OK. 25. Tona, K., V. Bruggeman, and O. Onagbesan. 2005. Day-old chick quality: Relationship to hatching egg quality, adequate incubation practice and prediction of broiler performance. Avian Poult. Biol. Rev. 16:109 119. 26. Wyatt, C. L., J. R. Weaver, and W. L. Beane. 1985. Influence of egg size, eggshell quality, and post hatch holding time on broiler performance. Poult. Sci. 64:2049 2055. 27. Vieira, S. L., and E. T. Moran Jr. 1998. Broiler yields using chicks from egg weight extremes and diverse strains. J. Appl. Poult. Res. 7:339 346. 28. Tona, K., O. Onagbesan, B. De Ketelaere, E. Decuypere, and V. Bruggeman. 2004. Effects of age of broiler breeders and egg storage on egg quality, hatchability, chick quality, chick weight, and chick post hatch growth to fortytwo days. J. Appl. Poult. Res. 13:10 18. 29. Fattori, T. R., H. R. Wilson, F. B. Mather, and S. M. Bootwalla. 1992. Strategies for weighing broilers, broiler breeder pullets and broiler breeder hens: 1. Sample size and individual vs. group weighings. J. Appl. Poult. Res. 1:88 94. 30. Gross, W. B., and P. B. Siegel. 1995. Factors which may affect the precision of experiments employing chickens. J. Appl. Poult. Res. 4:411 421.