The authors of ‘Limited filling of the potential range in European tree species’ (Svenning & Skov 2004), in the July 2004 issue of Ecology Letters, would like to point out an error in their paper, as detailed below.

While the GIS analyses should have included only European land areas, the mask used for projecting the bioclimatic envelopes also included the part of Asia Minor falling within the 34° N to 72° N and 11° W to 32° E study area boundaries. Redoing the analyses with Asia Minor excluded, produced median and mean R/P values of 29.7 and 39.8% (± 30.7% SD) instead of 27.7 and 38.3% (±30.3% SD), respectively. The number of species with R/P values ≥ 90% or unoccupied, but climatically suitable areas ≥ 20% or ≥ 1/3 of Europe, were unchanged. Both Pearson and Spearman rank correlations between the old and new R/P values were 0.999. Correspondingly, with respect to all the subsequent results, the correction did not change which relationships were significant or their sign, and had at most negligible effect on the exact values estimated.