APPENDICES

Appendix S1 Basic statistics on mean plant indicator values (MIV), landscape, climate, soil and forest stand variables over the 19989 sample plots. Summary statistics are means ± SD [min-max] for continuous variables and frequency in each class for nominal variables.

Variable name / Code / Type (1) / Summary statistics
MIV for pH (Ecoplant) / MIV.pH / C / 5.25 ± 1.07 [3.0-8.5]
MIV for N (Ellenberg) / MIV.N / C / 4.74 ± 0.69 [1-7]
MIV for L (Ellenberg) / MIV.L / C / 5.35 ± 0.54 [3-8]
MIV for F (Ellenberg) / MIV.F / C / 5.45 ± 0.38 [2-8.09]
MIV for Ta (Ecoplant) / MIV.Ta / C / 9.84 ± 1.03 [3.8-15.7]
Distance-to-edge (m) / DIST / C / 292 ± 254 [3-1096]
Forest patch-size (ha) / FPS / C / (2) 13.8·104 ± 26.8·104 [327-1 105]
(3) 1970 ± 4789
Mean annual max. temperature (°C) (4) / TMAX / C / 14.5 ± 1.0 [9.3-17.2]
Mean annual precipitation (mm) (4) / RAIN / C / 861 ± 200 [380-2356]
Soil water capacity (mm) / SWC / C / 95.8 ± 49.7 [0-180]
Topography / TOPO / N / Negative: lateral water inflow lower than outflow (n=1510)
Neutral: flat position (slope<5%) (n=16449)
Positive: lateral inflow higher than outflow (n=972)
Very positive: lateral inflow much higher than outflow (n=1058)
Soil type / SOIL / N / Calcarisols/Rendzinas (n=197)
Calcisols (n=904)
Rendzinas/Calcosols (n=377)
Brunisols/Colluviosols (n=3554)
Brunisols/Luvisols (n=582)
Acid luvisols/Waterlogged luvisols (n=608)
Acid brunisols (n=2403)
Ochreous brunisols (n=318)
Podzolic soils/Podzols/Waterlogged podzols (n=1441)
Arenosols/Rankosols/Lithosols (n=315)
Waterlogged brunisols (n=1931)
Gleysols/Stagnosols (n=191)
Pseudogleys (n=5520)
Planosols/Pelosols (n=520)
Depth of temporary waterlogging (cm) / DTW / N / 0-14 (n=1238)
15-34 (n=2240)
35-54 (n=2238)
55-74 (n=1049)
>75 (n=330)
Absence (n=12894)
Variable name / Code / Type (1) / Summary statistics
Depth of HCl effervescence (cm) / DHCl / N / 0-14 (n=1314)
15-34 (n=592)
35-54 (n=422)
55-74 (n=323)
>75 (n=233)
Absence (n=17105)
Canopy cover (%) / CCOV / C / 14.5 (n=39) ; 16 (n=242)
22 (n=69) ; 37 (n=660)
62 (n=1531) ; 75 (n=893)
87.5 (n=16555)
Tree species composition (5) / COMP / N / 1: deciduous forest (n=14969)
2: mixed stands (n=1550)
3: coniferous forest (n=3470)
Forest stand structure / STRUC / N / 1: even-aged high forest (n=11453)
2: high forest with coppice (n=7422)
3: coppice (n=1085)

(1) variable type: factor (F) or continuous variable (C). MIV are truncated continuous variables.

(2) mean patch size calculated at plot scale (19989 plots).

(3) mean patch size calculated at patch scale (1801 patches).

(4) climate data are normals over the 1971-2000 period.

(5) tree species composition: deciduous forest (canopy cover of deciduous trees 75-100%), mixed stands (deciduous or coniferous canopy cover 25-75%), coniferous forest (deciduous canopy cover 0-25%).

Appendix S2 Pearson correlation coefficient matrix for the 10 SITE predictors: for this analysis only, TOPO, DTW and DHCl were treated as continuous variables, DIST and FPS were log-transformed, and SOIL, COMP and STRUC were treated as factors. Relationships between factors were assessed with a Chi-squared test.

TOPO / DTW / DHCl / SWC / TMAX / RAIN / CCOV / DIST (1) / FPS (1) / SOIL / COMP / STRUC
TOPO / 1
DTW / 0.26 / 1
DHCl / -0.14 / -0.20 / 1
SWC / -0.19 / -0.49 / 0.18 / 1
TMAX / -0.22 / -0.19 / 0.07 / 0.04 / 1
RAIN / 0.27 / 0.23 / 0.04 / -0.12 / -0.50 / 1
CCOV / 0.03 / 0.00 / 0.03 / 0.04 / -0.05 / 0.04 / 1
DIST / -0.10 / -0.05 / 0.12 / 0.02 / -0.04 / 0.03 / 0.01 / 1
FPS / 0.11 / 0.08 / 0.07 / -0.13 / -0.20 / 0.29 / 0.03 / 0.14 / 1
SOIL / 0.30 / 0.64 / 0.76 / 0.66 / 0.26 / 0.33 / 0.09 / 0.14 / 0.17 / 1
COMP / 0.05 / 0.08 / 0.04 / 0.15 / 0.02 / 0.15 / 0.12 / 0.03 / 0.12 / 0.16 (2) / 1
STRUC / 0.03 / 0.02 / 0.05 / 0.04 / 0.12 / 0.09 / 0.20 / 0.10 / 0.08 / 0.17 (2) / 0.29 (2) / 1

(1) DIST and FPS were log-transformed.

(1) For SOIL, COMP and STRUC, we used Cramer's V statistics, a measure of association between two nominal variables, giving a value between 0 and +1 (inclusive), and based on Pearson's chi-squared statistic. The other nominal variables (TOPO, DTW and DHCl) were transformed into ordinal variable and considered as continuous variables.

Appendix S3 Variance partitioning of MIV for soil pH, N (soil nitrogen availability), L (light availability), F (soil humidity) and Ta (air temperature) according to distance-to-edge (DIST) and forest patch size (FPS). The table gives the pure effect of DIST, the pure effect of FPS, the joint effect of DIST and FPS, and the total variance explained (r²).

Pure effect of DIST / Pure effect of FPS / Joint effect of DIST and FPS / Total variance explained (r²)
MIV.pH / 0.040 / 0.051 / 0.005 / 0.095
MIV.N / 0.036 / 0.022 / 0.005 / 0.063
MIV.L / 0.002 / 0.004 / 0.001 / 0.006
MIV.F / 0.015 / 0.018 / 0.000 / 0.034
MIV.Ta / 0.004 / 0.094 / 0.004 / 0.102

Appendix S4 DEI assessment on MIV for soil pH, N (soil nitrogen availability) L (light availability), F (soil humidity) and Ta (air temperature). Differences in MIV between the deciles of the distribution of DIST were tested with ANOVA followed by a Tukey HSD test; means without common letters differ at P < 0.05.

9

Appendix S5 List of predictors in MSITE ranked in the order of selection in the stepwise multiple regression procedure; in brackets, the r² explained by each SITE predictor in a univariate regression.

List of predictors in MSITE
MIV.pH / SOIL (0.47), DHCl (0.20), COMP (0.08), RAIN-poly (0.03), SWC-poly (0.01), STRUC (0.07), TMAX-poly (0.01), DTW (0.02), TOPO (0.002), CCOV-poly (0.001)
MIV.N / SOIL (0.31), COMP (0.15), TMAX-poly (0.04), RAIN-poly (0.005), DHCl (0.06), SWC-poly (0.014), CCOV-poly (0.013), TOPO (0.006), DTW (0.007), STRUC (0.05)
MIV.L / COMP (0.17), TMAX-poly (0.11), CCOV-poly (0.12), SOIL (0.13), RAIN-poly (0.07), STRUC (0.04), DTW (0.02), SWC-poly (0.01), TOPO (0.006)
MIV.F / SOIL (0.29), DHCl (0.14), STRUC (0.06), COMP (0.04), TMAX-poly (0.001), SWC-poly (0.05), CCOV-poly (0.01), DTW (0.08), TOPO (0.006), RAIN-poly (0.01)
MIV.Ta / TMAX-poly (0.42), CCOV-poly (0.03), SOIL (0.07) TOPO (0.03), STRUC (0.02), COMP (0.001), SWC-poly (0.01), DTW (0.04), RAIN-poly (0.27), DHCl (0.003)

9

Appendix S6 Variance partitioning of MIV for soil pH, N (soil nitrogen availability), L (light availability), F (soil humidity) and Ta (air temperature) according to SITE, DIST and FPS.

Pure effect of SITE / Pure effect of DIST / Pure effect of FPS / Joint effect of SITE and DIST / Joint effect of SITE and FPS / Joint effect of DIST and FPS / Joint effect of SITE, DIST
and FPS / Total variance explained
MIV.pH / 0.486 / 0.016 / 0.008 / 0.026 / 0.053 / 0.002 / -0.007 / 0.581
MIV.N / 0.410 / 0.020 / 0.005 / 0.015 / -0.025 / 0.002 / 0.046 / 0.471
MIV.L / 0.445 / 0.001 / 0.002 / 0 / -0.002 / 0 / 0.003 / 0.450
MIV.F / 0.359 / 0.006 / 0.003 / 0.012 / 0.016 / 0 / 0.001 / 0.396
MIV.Ta / 0.425 / 0.002 / 0.011 / 0.003 / 0.065 / 0.001 / 0.021 / 0.529

9

Appendix S7 Variations in MIV for soil pH, N (soil nitrogen availability), L (light availability), F (soil humidity) and Ta (air temperature) according to distance-to-edge before (solid line) and after (dashed line) SITE correction. Curves come from locally-weighted regressions. MIV before SITE correction (i.e. MIV.pH) were centred for convenience. MIV after correction (i.e. MIV.pH.res) were residuals of the regression of MIV as a function of SITE.

Appendix S8 Results of multiple logistic regressions applied to each class of SITE factors using DIST and FPS as predictors. The table indicates the best model among the five models: null model, model using DIST, FPS, FPS and DIST, or FPS and DIST with interaction as predictor(s). The sub-total computes the occurrence of models as best model for SITE factors. See Appendix S1 for variable definition.

SITE factors / NULL / DIST / FPS / DIST + FPS / DIST x FPS
SOIL
Podzolic soils/Podzols/Wat. podzols / X
Ochreous brunisols / X
Planosols/Pelosols / X
Acid brunisols / X
Gleysols/Stagnosols / X
Waterlogged brunisols / X
Pseudogleys / X
Acid luvisols/Waterlogged luvisols / X
Arenosol/Ranker/Lithosol / X
Brunisols/Luvisols / X
Brunisols/Colluviosols / X
Rendzinas/Calcosols / X
Calcisols / X
Calcarisols/Rendzinas / X
Sub-total / 1 / 0 / 7 / 6 / 0
DHCl
0-14 / X
15-34 / X
35-54 / X
55-74 / X
>75 / X
absence / X
Sub-total / 0 / 0 / 4 / 2 / 0
TOPO
Negative / X
Neutral / X
Positive / X
Very positive / X
Sub-total / 0 / 1 / 0 / 3 / 0
SITE factors / NULL / DIST / FPS / DIST + FPS / DIST x FPS
DTW
0-14 / X
15-34 / X
35-54 / X
55-74 / X
>75 / X
absence / x
Sub-total / 0 / 0 / 4 / 2 / 0
TMAX / X
RAIN / X
SWC / X
CCOV / X
COMP
Deciduous / X
Mixed / X
Coniferous / X
Sub-total / 0 / 0 / 2 / 1 / 0
STRUC
Even-aged high forest / X
High forest with coppice / X
Coppice / X
Sub-total / 0 / 0 / 0 / 3 / 0

9