Fig. 4: The difference between five dominant and rare trait values and PC1 at the y-axis plotted against, respectively, mean annual temperature (left) and water availability (right).
From: Effect of climate on traits of dominant and rare tree species in the world’s forests
See for the other traits and PC2 Fig. S7. Positive values indicate that trait values are higher for dominant species, while negative values indicate that trait values are higher for rare species. The graphs on temperature are modelled with a second-order polynomial function, the graphs on availability index are modelled with a linear function as this best fitted the data (See Supplementary Data 1). The standardized regression coefficients of the linear models are displayed of the climatic variables (t = temperature, w = water availability index) and squared climatic variables (t2 = temperature2, w2 = water availability index2). The number of plots included in this analysis is 22,825. The significance of the regression coefficients is indicated with asterisks: *p < 0.05, **p < 0.01, ***p < 0.001. Source data is provided as a Source Data file.
