Maximum Ingredient Level Optimization Workbook for Estimating the Maximum Safe Levels of Feedstuffs (B 1469) University of Georgia Extension New feed ingredients are evaluated and introduced to the feed industry every year. The evaluation process is necessary and includes feeding birds different levels of the test ingredient to estimate the maximum safe level (MSL). The MSL is usually estimated with a multiple range test, ignoring the fact that this test is inappropriate for this type of feeding trials where the independent variable is continuous. This paper describes the use of the Maximum Ingredient Optimization Workbook (MIOW) in estimating the MSL and determining the optimal combination or ingredient levels and replications for most efficient experimental design of future feeding trials. The MIOW calculates the results and the related descriptive statistics (SD, SE, Cl, and R2) based on simulation and non-linear regression models (broken-line linear and broken-line quadratic models). 2017-02-03 11:53:22.657 2017-02-03 11:53:22.143 Maximum Ingredient Level Optimization Workbook for Estimating the Maximum Safe Levels of Feedstuffs | Publications | UGA Extension Skip to content

Maximum Ingredient Level Optimization Workbook for Estimating the Maximum Safe Levels of Feedstuffs (B 1469)

Download PDF

Summary

New feed ingredients are evaluated and introduced to the feed industry every year. The evaluation process is necessary and includes feeding birds different levels of the test ingredient to estimate the maximum safe level (MSL). The MSL is usually estimated with a multiple range test, ignoring the fact that this test is inappropriate for this type of feeding trials where the independent variable is continuous. This paper describes the use of the Maximum Ingredient Optimization Workbook (MIOW) in estimating the MSL and determining the optimal combination or ingredient levels and replications for most efficient experimental design of future feeding trials. The MIOW calculates the results and the related descriptive statistics (SD, SE, Cl, and R2) based on simulation and non-linear regression models (broken-line linear and broken-line quadratic models).

Click the "Download PDF" link to view this publication.

Status and Revision History
Published on Feb 3, 2017