Get access

Forecasting the future of biodiversity: a test of single- and multi-species models for ants in North America

Authors

  • Matthew C. Fitzpatrick,

  • Nathan J. Sanders,

  • Simon Ferrier,

  • John T. Longino,

  • Michael D. Weiser,

  • Rob Dunn


M. C. Fitzpatrick (mfitzpatrick@umces.edu), Univ. of Maryland Center for Environmental Science, Appalachian Lab, Frostburg, MD 21532, USA. – N. J. Sanders, Dept of Ecology and Evolutionary Biology, 569 Dabney Hall, Univ. of Tennessee, Knoxville, TN 37918, USA, and Center for Macroecology, Evolution and Climate, Dept of Biology, Univ. of Copenhagen, DK-2100 Copenhagen, Denmark. – S. Ferrier, CSIRO Ecosystem Sciences, Black Mountain Laboratories, Clunies Ross Street, Black Mountain ACT 2601, GPO Box 1700, Canberra ACT 2601, Australia. – J. T. Longino, The Evergreen State College, Olympia, WA 98505, USA. – M. D. Weiser, Dept of Biology, North Carolina State Univ., Raleigh, NC 27695, USA. – R. Dunn, Dept of Biology and Keck Center for Behavioral Biology, North Carolina State Univ., Raleigh, NC 27695, USA.

Abstract

The geographic distributions of many taxonomic groups remain mostly unknown, hindering attempts to investigate the response of the majority of species on Earth to climate change using species distributions models (SDMs). Multi-species models can incorporate data for rare or poorly-sampled species, but their application to forecasting climate change impacts on biodiversity has been limited. Here we compare forecasts of changes in patterns of ant biodiversity in North America derived from ensembles of single-species models to those from a multi-species modeling approach, Generalized Dissimilarity Modeling (GDM). We found that both single- and multi-species models forecasted large changes in ant community composition in relatively warm environments. GDM predicted higher turnover than SDMs and across a larger contiguous area, including the southern third of North America and notably Central America, where the proportion of ants with relatively small ranges is high and where data limitations are most likely to impede the application of SDMs. Differences between approaches were also influenced by assumptions regarding dispersal, with forecasts being more similar if no-dispersal was assumed. When full-dispersal was assumed, SDMs predicted higher turnover in southern Canada than did GDM. Taken together, our results suggest that 1) warm rather than cold regions potentially could experience the greatest changes in ant fauna under climate change and that 2) multi-species models may represent an important complement to SDMs, particularly in analyses involving large numbers of rare or poorly-sampled species. Comparisons of the ability of single- and multi-species models to predict observed changes in community composition are needed in order to draw definitive conclusions regarding their application to investigating climate change impacts on biodiversity.

Get access to the full text of this article

Ancillary