Model for predicting egg output and metabolisable energy intake of laying pullets.

1. The data compiled by Marsden and Morris (1987) to examine the relationships between environmental temperature and the long-term, adapted responses of laying pullets were divided at random into two subsets of 99 and 113 observations. The first subset was used to estimate regression coefficients for an econometric model, and the second subset to validate the model. 2. Equations to predict inputs (costs) and outputs (returns) were estimated with a three-stage least-squares regression model. Three stage least-squares estimation is a technique which corrects for the simultaneity of variables within the model and correlation across equations of the model. This results in more efficient estimates of the regression coefficients. 3. The final output and output equations were: MEI = 253.86-190.31EM+5.766EM2-0.546EM3 + 0.7034T-0.004388T3 + 695.08BW-120.23BW2 + 397.37ME-13.132ME2-1.06MEXT; R2 = 0.86; EO = 119 + 0.025MEI -0.0000045MEI2-1.462T-0.0791T2-135.3BW + 38.31BW2-1.483T X BW + 0.0288T2 X BW + 0.673 delta BW; R2 = 0.59 where MEI = daily metabolisable energy intake (kJ/bird d), T = environmental temperature (degree C), EO = egg output (g/bird d), BW = body weight, and ME = metabolisable energy concentration (kJ/g). The values for R2 indicate very good fits considering that the data were recorded over a 26-year period in 14 different laboratories. 4. This statistical model can serve as the basis for an econometric model of egg production to determine the environmental temperature that maximises profits from laying pullets of different body weights.