Fig 1.
A) Linear regressions of leaf mass per area (LMA) with leaf volume per area (LVA or thickness) and leaf density (LD), for deciduous species (dark line and circles), evergreens (red line and empty red circles) and all the species (dashed line). The level of significance is expressed as follows: * P < 0.05, ** P < 0.01, *** P < 0.001. Box plots (median and 1st and 3rd quartiles) of deciduous vs. evergreen species are included in the margins; the significant results are based on the phylogenetic generalised least squares model (PGLS). The results based on PGLS can be found in S3 Table, Supporting Information. Whiskers show the minimum and maximum values that fall within 1.5x the length of the box away from the interquartile range; data further away are shown as outliers. B) Decomposition of the total variability explained by LVA and LD.
Fig 2.
A) Linear regressions of leaf volume per area (LVA or thickness) with anatomical tissues, for deciduous species (dark line and circles), evergreens (red line and empty red circles) and all the species (dashed line). In brackets the slopes of the regression lines are given. The level of significance is expressed as follows: * P < 0.05, ** P < 0.01, *** P < 0.001. Box plots (median and 1st and 3rd quartiles) of deciduous vs. evergreen species are included in the margins; the significant results are based on the phylogenetic generalised least squares model (PGLS). The results based on PGLS can be found in S3 Table, Supporting Information. Whiskers show the minimum and maximum values that fall within 1.5x the length of the box away from the interquartile range; data further away are shown as outliers. B) Decomposition of the total variability explained by anatomical tissues, without covariations.
Fig 3.
A) Linear regressions of leaf mass area (LMA) and mesophyll with leaf carbon concentration (LCC) and leaf nitrogen concentration (LNC), for deciduous species (dark line and circles), evergreens (red line and empty red circles) and all the species (dashed line). In brackets the slopes of the regression lines are given. The level of significance is expressed as follows: * P < 0.05, ** P < 0.01, *** P < 0.001, + is shown when the relationship is phylogenetically dependent. Box plots (median and 1st and 3rd quartiles) of deciduous vs. evergreen species are included in the margins; the significant results are based on the phylogenetic generalised least squares model (PGLS). The results based on PGLS can be found in S3 Table, Supporting Information. Whiskers show the minimum and maximum values that fall within 1.5x the length of the box away from the interquartile range; data further away are shown as outliers.
Table 1.
Linear regressions of the morphological and anatomical traits with leaf carbon concentration (LCC) and leaf nitrogen concentration (LNC).
The positive or negative relations (Rel) and the R2 of the regressions are shown.
Fig 4.
Correlation network for morphological (white), anatomical (blue) and chemical traits (grey), describing the interrelations with LMA (yellow).
Blue lines indicate positive correlations and red lines negative correlations. Dashed lines indicate phylogenetic independence. The correlation coefficient is shown. Thin lines, 0.5<|r|<0.707 (0.25<R2<0.50); intermediate lines, 0.707<|r|< 0.866 (0.50<R2<0.75); bold lines, |r|>0.866 (R2>0.75).
Fig 5.
Linear regressions between soil water content and morphological traits (leaf mass per area, LMA and leaf density, LD) for deciduous species (dark line and circles), evergreens (red line and empty red circles) and all the species (dashed line).
The level of significance is expressed as follows: + 0.05 < P < 0.10, ** P < 0.01, *** P < 0.001.