Fig. 2: Distributional responses to long-term changes in land use and climate.

a The rate of mean temperature (T) change was calculated for Great Britain, at 10 × 10 km grid square resolution. b a ‘land conversion’ map was also generated, where values indicate the proportion of 25 × 25 m pixels within each 10 × 10 km square that changed category between the two time periods, interval ~75 years. Long-term persistences or extirpations for 1192 species were modelled in response to various combinations of land-use change, temperature change, and their interaction. Models fitted to the data for each taxon included both temperature change and land conversion, suggesting a positive effect of each variable on persistence (with an antagonistic interaction, in the case of birds), but a high proportion of these models’ explanatory power was accounted for by species-level random effects. Subsequent models fitted to data for each species revealed large variation in the Average Marginal Effects (AMEs) of (c) a 0.1 °C per decade temperature rise and a 10% land conversion (d). All 1192 species are illustrated in grey, with coloured bars representing the subsets of species for which the respective variable was included in the ‘best’ model (excluding species exhibiting interactive responses, n = 230).