Fangliang He and Stephen Hubbell correct an overestimation of 160% for species extinction rates resulting from habitat destruction (Nature 473, 368–371; 2011). However, near-term extinction rates predicted by the Millennium Ecosystem Assessment still remain at 400–4,000 times the background rate of species extinction.

Although it may help to refine future predictions, we caution against their recommendation for collating more detailed geographical data as an urgent priority for conservation science.

Knowing where species occur and their risk of extinction is fundamental for deciding where to focus efforts to protect them. But the diminishing returns on the value of biological surveys (H. S. Grantham et al. Conserv. Lett. 1, 190–198; 2008) means that more data may not translate into significantly better decisions. Heterogeneity in the costs and likelihood of success of conservation actions can influence investment priorities far more.

Areas designated a priority for species protection, identified using the 'species–area relationship', are not affected by model uncertainty, taxonomic group or the non-random distribution of species (M. C. Evans et al. Divers. Distrib. 17, 437–450; 2011).