Skip to main content

Table 5 Results of likelihood-ratio analysis of deviance of marginal mixed models using generalized least square with threatened species richness as a dependent variable

From: Threatened species richness along a Himalayan elevational gradient: quantifying the influences of human population density, range size, and geometric constraints

 

Df

AIC

LRT

P value

(Full model)

 

− 50.202

  

Area

1

− 44.958

7.244

< 0.01

Null model (simulated range)

1

− 51.967

0.235

0.627

Elevation

1

− 48.772

3.430

0.063

Empirical range: population density

1

− 48.078

4.124

0.042

  1. Likelihood-ratio analysis tested whether the deviance of the full model significantly increases after each explanatory variable was excluded from the model. The denominator Df = 28