Predicting range overlap in two closely related species of spiders

1. Predicting when and where species are likely to experience inter-specific interactions as a result of climate change may be as relevant to understanding their evolutionary futures as predicting responses to physical environmental variables. 2. In this paper, models built using data for the distributions of two species of large house spider, Tegenaria saeva and T. gigantea (Agelenidae), from relatively long-established parts of their ranges are used to predict species overlap in a region of more recent range expansion. 3. Generalised linear models (GLMs) are used to identify the key environmental variables associated with the distributions of the spider species in an east[long dash]west band across central Britain. Using this model, the distributions of the two species both north and south of the central band are predicted and tested for how closely they matched observed distributions, and thus whether environmental factors are a sufficient explanation for current distribution patterns in England and Wales. 4. Results demonstrate that predictions of both species distributions and overlap in the south region are almost as good as in the model-building zone, but that predicted distributions in the north are no better than random. Here either climate is of no importance in determining species distributions or, perhaps more likely, the system is still in a state of flux and currently reflects the stochasticity of recent colonisation. 5. Further long-term monitoring of the populations may allow discrimination between alternative hypotheses that could explain the current mismatch between climate and species' distributions