Table 8. Digestible Lys requirement from 28 to 42 d of age (Experiment 3).
| Model | Response variable | Equation | Estimated requirement (100, 99, 95%) | P-value | R2 |
|---|---|---|---|---|---|
| Quadratic polynomial1 | BW gain | Y = -1295.9 + 6102.9x – 3019.3x2 | 1.011, 1.001, 0.960 | 0.0001 | 0.8632 |
| FCR5 | Y = 4.605–5.629x + 2.727x2 | 1.032, 1.022, 0.981 | 0.0001 | 0.9500 | |
| Carcass yield | Y = 55.32 + 52.72x – 27.72x2 | 0.951, 0.941, 0.903 | 0.0001 | 0.3930 | |
| Breast meat yield | Y = -14.43 + 80.51x – 40.64x2 | 0.990, 0.981, 0.941 | 0.0001 | 0.7599 | |
| Linear broken-line2 | BW gain | Y = 1773.4–1572.7 × (0.879 –x) | 0.879, 0.870, 0.835 | 0.0001 | 0.8564 |
| FCR | Y = 1.712 + 1.437 × (0.902 –x) | 0.902, 0.893, 0.857 | 0.0001 | 0.9417 | |
| Carcass yield | Y = 80.27–9.13 × (0.878 –x) | 0.878, 0.869, 0.834 | 0.0001 | 0.3815 | |
| Breast meat yield | Y = 25.32–17.42 × (0.891 –x) | 0.891, 0.882, 0.846 | 0.0001 | 0.7713 | |
| Quadratic broken-line3 | BW gain | Y = 1777.3–3561.5 × (0.982 –x)2 | 0.982, 0.972, 0.933 | 0.0001 | 0.8664 |
| FCR | Y = 1.707 + 3.002 × (1.014 –x)2 | 1.014, 1.004, 0.963 | 0.0001 | 0.9511 | |
| Carcass yield | Y = 80.25–23.51 × (0.959 –x)2 | 0.959, 0.949, 0.911 | 0.0001 | 0.3627 | |
| Breast meat yield | Y = 25.30–43.83 × (0.974 –x)2 | 0.974, 0.964, 0.925 | 0.0001 | 0.7574 | |
| Exponential asymptotic4 | BW gain | Y = 1335 + 472 × (1 –EXP(– 7.544 × (X– 0.68))) | –,1.290, 1.077 | 0.0001 | 0.8588 |
| FCR | Y = 2.079–0.415 × (1 –EXP(– 6.433 × (X– 0.68))) | –,1.396, 1.146 | 0.0001 | 0.9456 | |
| Carcass yield | Y = 76.39 + 3.97 × (1 –EXP(– 8.499 × (X– 0.68))) | –,1.222, 1.032 | 0.0001 | 0.3332 | |
| Breast meat yield | Y = 19.88 + 5.75 × (1 –EXP(– 7.687 × (X– 0.68))) | –,1.279, 1.070 | 0.0001 | 0.7344 |
1 Quadratic polynomial regression model (QP): Y = β0 + β1 × X + β2 × X2, where Y is the dependent variable, X is the dietary Lys concentration, and β0 is the intercept, β1 and β2 are the linear and quadratic coefficients, respectively; maximum response concentration was obtained by:—β1 ÷ (2 × β2).
2 Linear broken-line model (LBL): Y = β0 + β1 × (β2—X), where (β2—X) = 0 for X > β2, Y is the dependent variable, X is the dietary Lys concentration, β0 is the value at the plateau, β1 is the slope and β2 is the Lys concentration at the break point.
3 Quadratic broken-line model (QBL): Y = β0 + β1 × (β2—X)2, where (β2—X) = 0 for X > β2, Y is the dependent variable, X is the dietary Lys concentration, β0 is the value at the plateau, β1 is the slope and β2 is the Lys concentration at the break point.
4 Exponential asymptotic (EA): Y = β0 + β1 × (1 –EXP(– β2 × (X– β3), where Y is the dependent variable, X is the dietary Lys concentration, β0 is the response for the dependent variable estimated for the feed with the lower Lys, β1 is the difference estimated between the minimum and maximum response obtained by the increasing Lys, β2 is the slope of the exponential curve, β3 is the Lys at the lower level; requirement were estimated by calculating (ln(0.05)/– β2) + β3 for 95% of the requirement and (ln(0.01)/– β2) + β3 for 99%.
5 Feed conversion ratio corrected for mortality weight.