Skip to main content

Table 2 Best subset of models explaining aboveground biomass productivity (AGBP)

From: Biomass and dominance of conservative species drive above-ground biomass productivity in a mediterranean-type forest of Chile

H

Models

Predictor variables

P

VIF

R 2 Adj

AICc

Ī”s

w

V-N

AGBPā€‰=ā€‰1.1TD ā€“ 0.4ln(S)

Model

<ā€‰0.001

Ā 

0.83

22.7

0

0.38

TD

<ā€‰0.001

1.7

Ā Ā Ā Ā 

ln(S)

0.04

1.7

Ā Ā Ā Ā 

M

AGBPā€‰=ā€‰0.88piC. alba

Model

<ā€‰0.001

ā€“

0.76

23.5

0.7

0.26

V

AGBPā€‰=ā€‰0.88TD

Model

<ā€‰0.001

ā€“

0.76

23.8

1

0.23

M-N

AGBPā€‰=ā€‰0.9piC. alba ā€“ 0.2ln(S)

Model

<ā€‰0.001

Ā 

0.77

26.6

3.8

0.06

p i C. alba

<ā€‰0.001

1.45

Ā Ā Ā Ā 

ln(S)

0.27

1.45

Ā Ā Ā Ā 

M-V

AGBPā€‰=ā€‰0.5piC. albaā€‰+ā€‰0.4TD

Model

<ā€‰0.001

Ā 

0.76

27.2

4.4

0.04

p i C. alba

0.34

10

Ā Ā Ā Ā 

TD

0.38

10

Ā Ā Ā Ā 

M-R

AGBPā€‰=ā€‰0.77TDā€‰+ā€‰0.17SD

Model

0.001

Ā 

0.75

27.5

4.7

0.04

TD

0.04

6.1

Ā Ā Ā Ā 

SD

0.97

6.1

Ā Ā Ā Ā 
  1. Legend: Hypotheses (H) tested, V Vegetation quantity; M Mass-ratio, N Niche complementarity, R Resources availability. Predictors, TD Tree density, pi species relative abundance, S Species richness, Hā€² Shannon index, SD Sand (%); R2Adjā€‰=ā€‰adjusted determination coefficient. All multiple regression models were statistically significant (Pā€‰<ā€‰0.05). The following acronyms were used: VIFā€‰=ā€‰Variance inflation factor; AICcā€‰=ā€‰Akaikeā€™s information criterion for small samples (AICc). Ī”sā€‰=ā€‰Difference among AICc of any model respect to the best. wā€‰=ā€‰Akaikeā€™s weight, it is the probability of a model to be the best in scale from 0 to 1