A Mechanistic dynamic simulation model of nutrient utilization, growth and body composition in pre-weaned lambs reared artificially

A mechanistic dynamic model was developed to simulate the growth and body composition of pre-weaned lambs reared artificially. The results of the simulated model were evaluated using data from the three experiments involving 79 twin-born male lambs randomly allocated to different combinations of milk replacer and pellet feeds resulting in ten treatment groups. The accuracy of simulation of the output traits was evaluated by mean bias, concordance correlation coefficient (CCC), relative predictive error (RPE) and the coefficient of determination (R-2). The model accurately simulated daily protein deposition (mean bias = 0.07 g /d +/- 1.5; R-2 = 0.93; RPE = 4.21 %; CCC = 0.96), water deposition (mean bias = -2.34 +/- 6.75 g /d +/- 1.5; R-2 = 0.97; RPE = 6.23 %; CCC = 0.94) and ash deposition (mean bias = 0.22 +/- 0.43 g /d +/- 1.5; R-2 = 0.83; RPE = 9.00 %; CCC = 0.87) rates. Fat deposition in lambs was underestimated (paired t-test, P < 0.05; mean bias = 3.75 +/- 4.99 g /d) due to an over-estimation of maintenance energy requirement for lambs given restricted access to pellet feeds and, therefore, resulting in an overall poor prediction of fat deposition rate (R-2 = 0.37 RPE = 29.78%; CCC = 0.34). Empty body weight at slaughter of the lambs was accurately simulated (R-2 = 0.95 RPE = 1.66%; CCC = 0.97). Although live weights (LW) and average daily gains (ADG) of the lambs were underestimated (paired t-test, P < 0.05), the accuracies of the overall simulated means for LW and ADG were substantial (R-2 = 0.90 and 0.99; RPE = 3.73 % and 6.11; CCC = 0.89 and 0.95 for LW and ADG, respectively). The model was re-run after adjusting for the maintenance energy function. The R-2 and CCC values for daily fat deposition increased from 0.37 to 0.46 and from 0.34 to 0.67 respectively, while the RPE decreased from 29.78% to 10.68% and overall actual and simulated mean fat deposited were not different (paired-t-test, P > 0.05). The evaluations presented indicated that the lamb growth model developed in this study can be used to simulate, with acceptable accuracy, pre-weaned lamb growth and body composition under different artificial rearing conditions where milk replacer, pellet intake and composition varies.