Methodology for partitioning poultry producers into recommendation domains

Abstract

This study simulated broiler grower conditions by incorporating environmental variation into the research design. This provided a means for partitioning growers into homogeneous production environments (recommendation domains). Also, it enabled a more accurate evaluation of experimental feed treatments and resulted in feeding recommendations appropriate to specific groups of broiler growers. Of all the variables in the environmental matrix, house temperature was the most important due to its effect on feed intake and thus weight gain. The simulation study was conducted in a broiler house with 32 pens (research domain). Each environment was a randomized complete block of four feed treatments. Broiler gain response to feed ingredients (milo and peanut meal) was used to evaluate the four isocaloric protein treatments fed in pellet form. Modified stability analysis was used to visualize the differences in the broiler gain response to treatments in ‘good’, ‘moderate’ and ‘poor’ environments, and to make feeding recommendations specific to environment or research domain. The ability to partition a broiler research domain was demonstrated and indicates that research conducted by simulating a well-characterized production system can be a practical approach to technology generation.