Table 5.
Multiple regression analysis to choose the most important CHO spectral parameters to predict CHO nutrient supply from the alfalfa forage (Medicago sativa spp. Sativa).
| Predicted Variables (Y) | Variable Selection (Variables Left in the Model with p < 0.05) | Prediction Equations (Test Model: Y = a + b1x1 + b2x2) | Model R2 Value | RSD a | p Value |
|---|---|---|---|---|---|
| Carbohydrate fraction profiles | |||||
| CA4 (% CHO) | Ratio of structural CHO to total CHO left in the model | Y = −41.04 + 99.32 A_StCHO/A_CHO ratio | 0.77 | 1.00 | 0.022 |
| CB1 (% CHO) | H_1415 left in the model | Y = −6.91 + 355.77 H_1415 | 0.70 | 0.97 | 0.038 |
| CB3 (% CHO) | H_895 left in the model | Y = 42.57 − 4338.00 H_895 | 0.91 | 0.64 | 0.004 |
| Total digestible nutrients | |||||
| tdNDF (% DM) | A_non-STCHO left in the model | Y = 21.84 − 83.78 A_non-STCHO | 0.88 | 0.28 | 0.006 |
a RSD: Residual standard deviation.