Skip to main content

Table 5 Summary of the selection process of the principal components and the fitting statistics of the forward stepwise regression

From: Alternative approaches to assessing the natural regeneration of Scots pine in a Mediterranean forest

Model

Fixed effects

Random Effect

Log-likelihood

Compared to model

Pr > (Chi)

(0)

Intercept

–

−543.56

–

–

(1)

Intercept

Plot

−512.74

(0)

<0.001

(2)

Intercept

Plot/transect

−512.40

(1)

0.412

(3)

PC1 com

Plot

−512.55

(1)

0.540

(4)

PC2 com

Plot

−508.86

(1)

0.005

(5)

PC2 com + PC1 ed

Plot

−508.20

(4)

0.252

(6)

PC2 com + PC2 ed

Plot

−507.81

(4)

0.147

(7)

PC2 com + altitude

Plot

−508.86

(4)

0.950

(8)

PC2 com + slope

Plot

−508.07

(4)

0.209

  1. The chosen model is in italics
  2. PC1 ed component related to the Ca and Mg, PC2 ed component related to the pH and the electric conductivity, PC1 com component related to the number of trees, PC2 com component related to the number of stumps