Table 5.

Prediction equations for energy concentrations in diets and validation of equations¹⁾,²⁾

Equation No.	n	Equation3)	RMSE	R2	Intercept4)		Slope5)
					Value	p-value	Value	p-value
18	232	DE = 965+0.72×GE−66.40×ADF	187	0.61	−13 (28)	0.646	−0.70 (0.11)	<0.001
19	262	ME = 0.98×DE−4.22×CP	55	1.00	31 (4)	<0.001	0.06 (0.02)	0.011
20	228	ME = 1,521+0.56×GE−65.30×ADF	184	0.59	21 (29)	0.470	−0.71 (0.11)	<0.001
21	185	ME = 3,847+38.69×EE−67.24×ADF	181	0.57	41 (30)	0.174	−0.77 (0.11)	<0.001
22	185	ME = 3,741+5.75×CP+39.84×EE−68.86×ADF	179	0.58	63 (29)	0.041	−0.76 (0.10)	<0.001

RMSE, root mean square of error; DE, digestible energy; GE, gross energy; ADF, acid detergent fiber; ME, metabolizable energy; CP, crude protein; EE, ether extract.

¹⁾Thirty two in vivo data points were employed to validate prediction equations for energy concentrations in feed ingredients.

²⁾Values in parentheses are standard error.

³⁾Energy concentrations and chemical composition are expressed as kcal/kg dry matter and % dry matter, respectively.

⁴⁾The intercept represents the mean bias.

⁵⁾The slope represents the linear bias.