NeoBiota (Apr 2013)

Correlations between global and regional measures of invasiveness vary with region size

  • Wayne Dawson,
  • Lidewij Keser,
  • Marten Winter,
  • Petr Pyšek,
  • John Kartesz,
  • Misako Nishino,
  • Nicol Fuentes,
  • Milan Chytrý,
  • Laura Celesti-Grapow,
  • Mark van Kleunen

DOI
https://doi.org/10.3897/neobiota.16.4351
Journal volume & issue
Vol. 16, no. 0
pp. 59 – 80

Abstract

Read online

We aimed to assess the utility of the Global Compendium of Weeds (GCW) as an indicator of plant invasiveness, by relating it to invasiveness at smaller scales. We correlated two global measures of invasiveness for alien plant species taken from the GCW (the total number of references for each species and the number of continental areas they are reported from), against distribution data from 18 regions (countries and continents). To investigate relationships between correlation strength and region size and spatial resolution (size of distribution units), we conducted meta-analyses. Finally, invasiveness measures were correlated against the number of habitats occupied by alien plant species and their median abundance in those habitats, in fine-scale vegetation plots in the Czech Republic and the state of Montana (USA). The majority of Spearman’s rho coefficients between GCW-derived invasiveness and regional distributions were less than 0.4. Correlation strength was positively related to region size and resolution. Correlations were weaker when the number of habitats occupied by a species, and species abundances within occupied habitats, were considered. We suggest that the use of the GCW as an invasiveness measure is most appropriate for hypotheses posed at coarse, large scales. An exhaustive synthesis of existing regional distributions should provide a more accurate index of the global invasiveness of species.