Data extraction and collection
For each relationship reported in the studies, we gathered information
on 17 moderators potentially responsible for differences in the
dispersal - range size relationship across studies (Table 1). We
classified these moderators into six groups of variables related to: 1)
departure/transfer stage of dispersal, 2) settlement stage of dispersal,
3) evolution, 4) dispersal approximation, 5) range size definitions, and
6) taxonomic delimitation (see Table 1 for the description of the
variables and predictions).
To assess which traits may be most suitable to approach the dispersal
process and are most relevant for increasing range size, we reported the
effect of each dispersal-related trait included in each model examining
the dispersal-range size relationship (Table 2). We classified this
trait overview by realm (terrestrial vs. marine) and taxon (clade), due
to potential differences between these systems with respect to
dispersal-limiting factors (Table 1) and thus their relationship of
dispersal-related traits to range size (Table 2). To examine the effect
of the moderators on the dispersal-range size relationship for each
relationship, we reported the overall effect of dispersal on range size
(see Table S5 in Database). When the study reported two or more
dispersal-related traits with opposite effects on range size, the
overall effect was treated as ‘neutral’. When the study included several
dispersal-related traits, some with neutral effects and some with
positive effects, we treated the overall effect as ‘positive’, while we
treated the overall effect as ‘negative’ when the opposite happened.
For the calculation of biogeographical region size (which may influence
the opportunity for range expansion due to available area) we used GIS
layers (Spalding et al. 2007, Olson et al. 2004, and Natural Earth).
Among the methodological features, the number of dispersal-related
traits refers to the number of traits included as independent factors
(e.g., body size, diaspore size, plant height) or combined in a single
factor as a complex metric of dispersal (e.g., hand-wing index, PCA
axis, relative measurements). For instance, hand-wing index is used as a
single factor that is composed of two dispersal-related traits (wing
length and first-secondary length). Taxonomic breadth, the lowest
taxonomic level that includes all species in the study, was assigned
based on the species lists provided in the studies. Relationships using
‘family’ and ‘superfamily’ taxonomic levels were recoded as ‘family’.
Similarly, ‘order’ and ‘suborder’ were recoded as ‘order’, and ‘phylum’,
‘division’, ‘subphylum’ and ‘subkingdom’ were recoded as
‘phylum-division’ level. 13 relationships could not be assigned to a
particular taxonomic breadth, thus excluded from statistical analyses.
Table 2. Summary of the observed relationship between
dispersal-related traits and range size based on 86 studies and 478
relationships. The relative proportion of the direction of the
relationships is indicated (positive: green, neutral: brown, negative:
red) for each realm (marine, terrestrial, freshwater) and taxon or clade
(fish, echinoderms, crustaceans, molluscs, amphibians, birds, mammals,
insects, vascular plants, liverworts, mycorrhiza, testudines,
trematodes, diatoms). The number of relationships assessed for each
taxon is indicated with N . In case of categorical variables, the
direction of the trait states is indicated below the trait, e.g.,
‘pelagic > non-pelagic’ means that species with pelagic
eggs have larger range sizes than those with non-pelagic eggs, and this
directional relationship was found in 2 out of 3 cases.