<div>*NA is reported in correspondence of the GLM predictor (Binary sig.) as
no analysis of variance had been computed for such categorical variable.</div><div>Figure 1. Map of the sampling sites of specimens considered for this
study across the Central Europe.</div><div>Figure 2. Flow chart illustrating the two alternative analyses to the
Mantel test used to detect the intraspecific relationship between
geographical distances and genetic distances. Left side: PCA method;
Middle: the NMDS method; it starts with the same matrix and ends with a
set of axes representing the DNA distances. Right side: PCNM technique
performed on geographic coordinates. The PCNM produced a set of
eigenvectors which had been regressed in Procrustes superimposition
analysis first against NMDS axis, then against PCA eigenvectors.
Finally, significance test with Protest.</div><div>Figure 3. Mean intraspecific genetic distances plotted vs A) mean
geographic intraspecific distances (in km); B) number of sampled
individuals per species (N_ind); C) number of sampling sites per
species (loc); D) vs geographic distances (km) for guilds with different
feeding styles; E) vs geographic distances (km) for guilds with
different habitat preferences; F) vs geographic distances (km) for body
size classes; G) vs geographic distances (km) for guilds of biotope
preference.</div><div><b>Figure 4</b> . Violin plots and bar plots of statistic scores from
Mantel, NMDS+PCA, and PCNM+PCA analyses across the different ecological
guilds and subcategories, differentiated for species with significant
(dark grey) and non- significant (light grey) correlation.</div><div>Habitat preferences: h_b =soil, h_e=eurytop, h_f= rotting matters,
h_n=nest, h_p=vegetation, h_t=dead wood, h_v=vegetation; h_w=
water; biotope preferences: b_e= eurytopic; b_f =wetlands ; b_o=
open-land biotypes; b_w=forests; body size classes: s_xs= extra small;
s_s= small; s_m=medium; s_l=large and s_xl= extra-large; feeding
style: f_c = coprophagous; f_e = polyphagous; f_m= mycetophagous;
f_n= necrophagous;f_p= phytophagous; f_s= saprophagous; f_x=
xylophagous; f_z= zoophagous; geographical distance classes: highest
treashold measured in km.</div><div>Supplement Figure 1. Distribution of the number of specimens (above) and
species (middle) on the five classes of geographical distances and the
amount of sampling localities (below).</div><div>Supplement Figure 2. Bar plots (above) showing the number of species
across the ecological guilds. Violin plots (below) illustrate the number
of specimens per species for each considered ecological trait: habitat
preference (A); feeding style (B); biotope preference (C); body size
classes (D); the width of violin indicates the total number of such
cases.</div><div>Supplement Figure 3. Relationships between pairwise geographic and
genetic distances across guilds for each ecological variable.</div><div>Supplement Figure 4. Dot plot showing the relationship between the
sampling variables: number of localities (loc); number of individuals
per species (N_ind); mean intraspecific geographic distance (km).</div><div>Supplement Figure 5. Significance levels for Mantel, NMDS, and PCA tests
(y-axis: significance value; x-axis: p-value). On the left panel,
significant species are shown in detail; right panel: all the examined
species that showed a positive and significant relationship for at least
one test.</div><div>Supplement Figure 6. Comparison between the three correlation methods.
on the bottom-left are regressed the three technique Statistic scores.
On the top-left are regressed the significance scores.</div><div>Supplement Figure 7. Histogram showing the 1785 species distribution
across the 95 examined families. In the chart all the species resulting
significant in at least one of the three statistical tests performed
during the study have been filled in dark grey.</div>