Determinants and Congruence of Species Richness Patterns across Multiple Taxonomic Groups on a Regional Scale
Table 1
Variables used for the multiple regression modelling and multiple regression models established for species richness of different taxa. Listed are the beta coefficients and significance levels for the parameters in the final models established for five taxonomic groups. The explained variance in species richness for each taxonomic group is also shown. For spatial validation of models, the Spearman rank correlation coefficients of predicted and observed species richness values and the respective significance level are shown. *, **, ***, n.s.: nonsignificant parameters in the final model that improve the explanatory power of the model (AIC).
Mean temperature of the coldest month (January) (°C)
TMPJAN
—
—
1.269**
—
−0.064n.s.
Spatial autocorrelation
SAC
0.053**
0.159***
0.409***
0.078***
0.005***
Number of plant species
not tested
not tested
0.071***
—
not tested
Model accuracy and validation
Goodness of fit (deviance change)
38.36%
33.76%
72.73%
62.34%
72.6%
Deviance change explained by environment
44.34%
—
45.42%
17.45%
7.05%
Deviance change explained by SAC
55.66%
100%
54.48%
82.55%
92.95%
ANOVA (versus Null model)
***
***
***
***
***
Spatial validation of the SAC model
0.758***
0.646***
0.752***
0.648***
0.721***
Spatial validation of the model without the SAC term)
0.606***
—
0.494***
0.326***
0.600***
a mean temperature in July/mean annual sum of precipitation.
bSimpson’s index of diversity using the relative proportion of the land-cover classes in grid cells.