Skip to main content
. 2022 Jun 17;13:895720. doi: 10.3389/fpls.2022.895720

Table 1.

Results of multiple linear regression (OLS) models for all environmental variables and the proportion of total, woody, and herbaceous samara species to total species.

t VIF Dev, % AIC Moran’s I
All plants
Sqrt(AP) 14.66*** 3.43 - - -
TS 16.44*** 2.05 - - -
MAT 9.99*** 2.26
MTCQ 5.42*** 1.73 - - -
Sqrt(PDQ) −5.07*** 3.17 - - -
ARI 3.22** 1.52 - - -
- - 46.91 1,271 0.219***
Woody species
Sqrt(AP) 18.75*** 3.37 - - -
TS 18.51*** 2.54 - - -
Sqrt(PDQ) −10.21*** 3.05
MTCQ 8.19*** 1.68
ARI 2.43** 1.41 - - -
- - 40.67 1,275 0.312***
Herbaceous species
MAT 8.74*** 2.36 - - -
TS 9.92*** 1.63 - - -
MDR −8.18*** 3.66 - - -
Sqrt(AP) −3.97*** 3.65 - - -
MTCQ 3.11** 1.59 - - -
- - 34.03 1,095 0.098***

***p < 0.001, **0.001 < p < 0.01, *0.01 < p < 0.05.

AP, mean annual precipitation; MAT, mean annual temperature; TS, temperature seasonality; MTCQ, the temperature of the coldest quarter; MDR, mean diurnal range; ARI, relief degree of land surface; PDQ, precipitation of the driest quarter; unique-R2, differences between the R2 of full MLR models and that of MLR models without that predictor; VIF, variance inflation factor, used to evaluate the significance of multi-collinearity; Dev, percentage deviance explained by the models; ‘-’, no value. If VIF is greater than five, then multi-collinearity is considered significant. Significance was determined using Student’s t-test (α = 0.05).