Table 1. Best-fit mixed-effects logistic growth model selection results and parameter estimates for canopy shrub volume, shrub density and ecological functioning.

Best model parameters					Parameter estimates
Fixed effects		Random effects	K	Δ AIC 1	Sample period	Coefficient			Δ AIC 2
						Asym	Scale	Xmid
Shrub volume
	Asym + Xmid + Scale	Asym, Scale	8	−5.67	1977	5.53	222.02	489.75	15.69
					1981	5.86	265.02	489.75	6.38
					1989	5.22	181.61	489.75	15.69
					2008	4.27	60.31	489.75	−1.97
Shrub density
	Asym + Xmid + Scale	Asym	6	−5.14	1977	0.92	511.30	361.95	−2.67	**
					1981	0.74	511.30	361.95	9.31
					1989	0.69	511.30	361.95	9.26
					2008	0.69	511.30	361.95	9.31
Run-on:Run-off area
	Asym + Xmid + Scale	Asym, Xmid, Scale			2008	4.43	470.00	838.33
Proportion of intact mounds
	Asym + Xmid + Scale	Asym, Xmid, Scale			2008	0.91	59.34	312.44

Coefficients vary by Sample Period where they differ significantly from population coefficients, while non-significant coefficients are represented only by the population value. Coefficients were considered significantly different (p<0.05) from zero.

Asym, Asymptote; Xmid, Curve inflection point; Scale, Inverse of curve steepness.

K, Number of model parameters; AIC, Akaike information criterion; ΔAIC ₁, AIC difference between the full model with random effects for each Sample Period associated with all fixed parameters and the best model with a reduced random effects structure; ΔAIC ₂, AIC difference between a model with separate parameters for each Sample Period and a model with separate parameters for the selected period only;

^**Sample Period different from all other periods combined.