Stepwise regression analysis for body weight at marketing age in different types of fowls

Abstract EN

Data of shank length (SL) in ‘mm’ and body weight (BW) in grams were collected for six groups of fowls, were analyzed to rank the SL and BW measurements at different ages bi-weekly up to marketing age (12 weeks of age for chickens, ten weeks for ducklings and 6 weeks of age for quail) according to their importance in influencing BW at marketing age.

The following two criteria were used to compare the relative efficiency of various models and to select the most suitable model, the coefficient of determination (R2) and the standard error of the estimate (SEE).

A larger value of coefficient of determination (R2) and smaller value of standard error of the estimate (SEE) indicate best fit of the model. The following results were obtained: The most important predictors (P<0.001) for chickens’ BW12 were SL12, BW4, BW0 and SL8 .Four predictors: BW6, SL10, BW4 and BW8 were the most important factors in determining ducklings’BW10, whereas BW5, SL6 and BW4 were significantly (P<0.01) affected Japanese quail’ BW 6.

Except for SL4, each of BW8, SL12, BW4 ,BW0 increased BW at marketing age for chicken males or females, shorter SL4 increased BW12 of both chicken’ males or females.

SL12 followed by BW8 and BW4 were the most important predictors for PP (a line of Fayoumi selected for high egg production in the first 90 days of production after five generations of selection) females’ BW12 with higher R2 and lower SEE than females of Dandarawi (Dand) and Golden Montazah (GM).

The first important trait in predicting BW10 for Pekin males was BW8 ,however BW6 and SL4 were the important predictors for Muscovy males’ BW10.

Either BW or SL at eight weeks of age were the most important predictors in determining Pekin females’BW10.

The first determining factor for predicting Pekin’ body weight at marketing age was BW8.