Skip to main content
ARS Home » Research » Publications at this Location » Publication #141881


item Cason Jr, John
item Young, Louis
item Berrang, Mark

Submitted to: European Symposium on Quality of Poultry Meat
Publication Type: Abstract Only
Publication Acceptance Date: 5/15/2003
Publication Date: 9/23/2003
Citation: Cason Jr, J.A., Young, L.L., Berrang, M.E. 2003. Use of paired chicken carcass halves in testing of antibacterial treatments [abstract]. European Symposium on Quality of Poultry Meat. 1:242.

Interpretive Summary:

Technical Abstract: Computer-generated data sets were created from previously collected rinse samples to test the relative efficiency of a random design based on counts of bacteria on whole chicken carcasses versus a random block design using counts from paired carcass halves for testing the effects of different processing treatments on carcass bacteria. From rinse samples of 45 pre-chill processed chicken carcasses, bacterial counts from 10 whole carcasses were randomly assigned to a control group and 10 to a treated group, or 10 carcasses were randomly selected and the right and left sides of each carcass were randomly assigned to either the control or treated groups. The process was repeated until 50 different data sets had been generated for each design. Mean bacterial counts for the original 45 carcasses were 3.0, 1.9, and 1.4 log10(cfu/ml) for aerobic bacteria, coliform, and E. coli counts, respectively. Means for each of the 50 generated data sets agreed within 0.1 log1 cfu/ml) for each kind of bacteria, indicating that the randomization was unbiased. Analysis of variance was conducted on each of the data sets and 0.1 was subtracted successively from counts in the treated group until a significant difference was obtained (P < 0.05). Mean reductions in counts necessary to obtain a significant difference between control and treated groups for aerobic bacteria, coliforms, and E. coli were 0.44, 0.51, and 0.63 for the random design and 0.25, 0.26, and 0.35 for the random block design respectively, with significant differences between the two designs for all three kinds of bacteria. If experimental conditions allow the use of half carcasses instead of whole carcasses in testing effects of treatments on carcass bacteria, a block design based on paired carcass halves is more sensitive to treatment differences.