Table 1 Percent contribution, permutation importance and gain for the environmental variables included in the model.

From: Prediction scenarios of past, present, and future environmental suitability for the Mediterranean species Arbutus unedo L.

Variable

Percent contribution

Permutation importance

Training gain without

Training gain with only

Mean temperature of driest quarter (BIO9)

28.3 (1)

2.1 (10)

99.8% (12)

58.9% (2)

Annual mean temperature (BIO1)

21.1 (3)

43.0 (1)

94.9% (7)

71.1% (1)

Slope

19.9 (2)

5.5 (8)

95% (1)

34.4% (4)

Temperature annual range (BIO7)

9.3 (4)

9.2 (5)

97.9% (6)

27.6% (7)

Precipitation of wettest quarter (BIO16)

7.3 (5)

17.3 (2)

96.4% (2)

9.4% (12)

Precipitation of warmest quarter (BIO18)

7.0 (10)

7.3 (3)

98.4% (4)

43.1% (3)

Precipitation seasonality (BIO15)

4.4 (7)

11.4 (4)

96.9% (5)

16.1% (10)

Isothermality (BIO3)

2.7 (9)

4.2 (7)

97.2% (3)

30.6% (5)

  1. In parenthesis the rank of the variable contribution/importance in the model fitted with the whole set of 12 variables. Training gain values correspond to ratios relative to the full model.