• Bioclimatic models;
  • conservation;
  • dispersal;
  • extinction;
  • fynbos;
  • GAM;
  • global change


Modelling of climate change-induced species range shifts has generally addressed migration limitations inadequately, often assuming ‘null’ migration or instantaneous ‘full’ migration extremes. We describe methods for incorporating simple migration rate assumptions into multispecies modelling, using the Proteaceae of the Cape Floristic Region. Even with optimistic migration assumptions, range loss projections more closely approximate null migration than full migration assumptions. Full migration results were positively skewed by few species with large range increases, an overestimate eliminated by dispersal-limited migration rate assumptions. Wind- and ant/rodent-dispersed species responded differently to climate change. Initially larger ranges of wind-dispersed species were more strongly reduced by climate change, despite far greater assumed dispersal distances — we suggest that these well-dispersed species populate more marginal areas of potential range, causing lower resilience to climatic changes at range margins. Overall, range loss rate slowed with advancing climate change, possibly because species ranges contracted into core areas most resilient to climate change. Thus, a consideration of simple dynamics of range change (rather than single step, present–future comparisons of range) provide new insights relevant for conservation strategies, in particular, and for guiding monitoring efforts to detect and gauge the impacts of climate change on natural populations.