Table 5.

The final mixed-model regression equations

	Fixed effectsb								R 2
Itema	Intercept	SE	Linear	SE	Quadratic	SE	Cubic	SE	Marginalc	Conditionald
Forage mass
April	1,748.5**	166.7							0.04	0.63
July	1,453.0**	181.7	177.2**	49.4					0.12	0.59
September	927.8**	311.2	NS	—	319.4**	99.6	−61.7**	21.3	0.14	0.67
Forage allowance
April	−3.34	3.43	8.93**	0.62					0.59	0.85
July	−3.24	2.22	7.11**	0.41					0.62	0.84
September	−1.18	2.33	NS	—	3.81**	0.40	−0.58**	0.09	0.68	0.87
Grazing days, AUD/ha	178.7**	19.4	−104.5**	21.3	26.4**	7.5	−2.4**	0.8	0.83	0.99
BW
Initial	215.9**	4.8							<0.01	0.96
Mid-summer	292.9**	9.8	4.5**	1.0					0.02	0.95
Ending	320.9**	13.0	8.3**	1.2					0.04	0.97
Average daily gain
Early summer	0.9**	0.06	0.04**	0.01					0.05	0.85
Late summer	0.5**	0.10	NS	—	0.01**	0.002			0.03	0.95
Total season	0.66**	0.04	0.05**	0.01					0.11	0.88
Gain per hectare	132.0**	9.9	−46.1**	6.7	5.4**	1.1			0.58	0.93
Net return with land cost
Per steer	120.2**	12.1	−15.8**	2.3					0.16	0.85
Per hectare	74.9**	5.6	−15.7**	1.4					0.49	0.84
Net return without land cost
Per steer	492.0**	146.3	−400.0*	160.9	144.7*	56.6	−16.2*	6.4	0.11	0.86
Per hectare	101.8**	5.6	−15.6**	1.4					0.49	0.84

The models implemented linear, quadratic, and cubic effects of stocking rate (hectares/steer) across the dependent variables. Model terms were only included in the regression when they were found to be significantly (P < 0.05) different from zero.

^aThe dependent variables were forage mass, kg forage dry matter (DM)/ha; forage allowance, kg forage DM/kg steer BW; BW, steer BW, kg; average daily gain (ADG), kg/d; net return with land cost, net return in USD on per steer or per hectare basis including a $ (USD) 27.17/ha land rental cost; Net return without land cost, net return in USD on per steer or per hectare basis without cost of land included in the analysis.

^bThe regression analysis explores the effects of stocking rate, with linear, quadratic, and cubic terms, on the dependent variables. Cubic, quadratic, and linear fixed effects were explored depending on the significance of the corresponding polynomial contrasts.

^cProportion of the variance of each dependent variable explained by the fixed effect of treatment only.

^dThe proportion of the variance which was not explained by the fixed effects but was explained by the random effect (year) is also reported, with the remainder of the variance being the residual standard deviation.

*Indicates significance at 0.01 < P ≤ 0.05 and **indicates P ≤ 0.05. In the event that a higher order term was significant but a lower order term was not, NS was used to specify nonsignificant in the place of the coefficient.