Comparison of Different Anticoccidials on Processing Characteristics and Parts Yield of Broiler Chickens

Abstract

Broilers were grown to 42 days of age on diets supplemented with salinomycin (60 mg/kg), monensin (99 mg/kg), or halofuginone (3 mg/kg) and continued on unmedicated diets to 49 days of age. There were no significant (P greater than .05) differences among anticoccidials in final body weight, feed conversion, or mortality rates. Samples of birds were processed for dressing percentage and parts yield. Both males and females fed salinomycin had significantly higher breast meat yield as a percentage of postchill weight than those fed halofuginone but not those fed monensin; differences were not significant for breast meat yield of males or females fed monensin or halofuginone. Males fed halofuginone had significantly heavier leg quarters than those fed salinomycin but not those fed monensin. Females fed salinomycin had significantly higher water uptake during chill than those fed monensin or halofuginone. Results of the present study indicate that the anticoccidial used in growing broilers may influence some carcass yield parameters.

Searching for Significance: The Perils of Excluding Pseudo-Outliers

“…In Experiment 3, the null hypothesis is true and the rejection rates should equal the theoretical probability of a Type I error (the significance level chosen by the researcher). As shown in the table, the test procedure of Izat et al (1991) leads to rejection rates that are three to six times higher than the theoretical significance level of the test (the rejection rate the test is supposed to produce). For example, this could lead a researcher using such a truncated sampling technique to claim a P = .05 when the actual value was closer to P = .19.…”

“…The rejection rates for the null hypothesis of the correct test and the test that duplicates the method used by Izat et al (1991) are displayed in Table 2. In Experiment 3, the null hypothesis is true and the rejection rates should equal the theoretical probability of a Type I error (the significance level chosen by the researcher).…”

“…The techniques reported by Ajuyah et al (1991), Halvorson et al (1991), and Izat et al (1991) were selected solely because they were recent papers using commonly reported methods of selecting data. The purpose of this work was not to criticize these authors' works or to question the validity of their data or conclusions, but rather to use their techniques to illustrate potential shortcomings of such practices.…”

“…A third experiment was conducted attempting to duplicate the procedure in Izat et al (1991) as closely as possible. Data were generated from the normal distribution using the mean and SD of Experiment 1 of Fletcher and Cason (1991) for 24 pens (12 pairs) with 50 birds to a pen and no treatment effect (all birds have the same expected weight).…”

“…Several recent studies published in Poultry Science (Ajuyah el ah, 1991;Halvorson et al, 1991;Izat et al, 1991; among numerous others dated back several years) have employed sampling strategies designed to include only birds near the pen mean to statistically test hypotheses concerning the effect of a specified treatment on some characteristic of interest (e.g., parts yield). The identification of outliers can be a valuable procedure for achieving accurate test results; a large literature exists on methods of detecting outliers (Rousseeuw and van Zomeren, 1990).…”