A influência da parcela e das repetições sobre as medidas de dispersão dos dados de desempenho de frangos de corte

Abstract

The objective of this study was to evaluate the influence of plot size and the number of replicates on feed intake, body weight, feed conversion, carcass yield and prime chicken cuts. In this experiment, 1050 one-day-old Cobb500® chicks were housed in experimental shed with 50 identical boxes with 10 square meters. The standard density was 12 birds/m2, so that the retaining ring could be adjusted according to the number of animals in proportion to maintain the same density. Two diets were offered ad libitum (initial ration of 1 to 21 days of age and other growth during the period 22 to 38 days old). In order to collect the performance data, the feed and the birds were weighed weekly throughout the experimental period and carcass yield and prime cuts were used two birds each, slaughtered after fasting for eight hours and then weighted. For each of the five planned plot sizes (5, 10, 20, 30, 40 birds) were simulated experimental plans, in a randomized design for repetitions (5, 8 and 10) totaling 15 scenarios. For each simulated experiment analysis of variance was performed. It was calculated the overall average and the coefficient of variation for each variable of the data collected and estimated the variation index proposed by GOMES (1991) for performance and carcass yield and cuts. The results suggested that for the performance study is recommended the use of plots with 30 animals and 5 repetitions and carcass yield and prime cuts 20 animals and 5 repetitions. The variation index was found to be an advantageoussubstitute for the coefficient of variation.