Table 5.

Statistical significance of five linear and linear plateau models used to determine the minimum effective and lowest maximum effective lubabegron (LUB) doses for reducing ammonia (NH₃) gas emissions per kilogram of body weight and hot carcass weight over the entire 91-d period¹

	Model2
Variable	1	2	3	4	5
NH3/kg BW, P-values	0.0107	0.0005	0.0004	0.0033	0.0052
NH3/kg HCW, P-values	0.0056	0.0002	0.0001	0.0015	0.0025

¹The minimum effective dose for NH₃ gas emissions per kg BW and HCW was determined to be the smallest dose used in the study that differed from the control based on planned contrasts performed following a significant F-test (P < 0.05). To determine the lowest maximum effective dose, a dose–response curve fit to the least squares means of the doses was performed. If the dose–response curve was determined to be a linear plateau model (Anderson and Nelson, 1975) and the slope or slopes were different (P ≤ 0.05) from 0, then the maximum effective dosage was the “join point” where the plateau began.

²Five competing linear and linear plateau models were evaluated based on the smallest P-value indicating best fit: (i) Linear = linear from 0 to 22.0 mg·kg⁻¹ DM; (ii) Quadratic1 = linear from 0 to 1.38 mg·kg⁻¹ DM, plateau from 1.38 to 22.0 mg·kg⁻¹ DM; (iii) Quadratic2 = linear from 0 to 5.5 mg·kg⁻¹ DM, plateau from 5.5 to 22.0 mg·kg⁻¹ DM; (iv) Quadratic3 = no response from 0 to 1.38 mg·kg⁻¹ DM, linear from 1.38 mg·kg⁻¹ DM to 5.5 mg·kg⁻¹ DM, plateau from 5.5 to 22.0 mg·kg⁻¹ DM; and (v) Quadratic4 = no response from 0 to 1.38 mg·kg⁻¹ DM, linear from 1.38 to 22.0 mg·kg⁻¹ DM.