Referee 2's review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

Review of Loyola et al. "Hung out to dry: choice of ecoregions for conserving threatened Neotropical anurans depends on life-history traits" for PLoS ONE

The authors address the interesting question of the interaction between life-history traits and priority regions for conservation, for the case of threatened Neotropical amphibians. Their results indicate that priority areas for species with aquatic larvae (AL) differ to some degree than those for species with terrestrial development (TD). I have three high-level concerns about the manuscript, concerning underlying objectives, data, and analytical techniques:

1) Underlying objectives. The authors' analyses target the representation of 95% of threatened amphibians, and then 95% of AL plus 95% of TD species. The rationale for these targets is not stated, and not at all obvious. Why would one want to set a target of 95%? It seems to me that if one is wanting to conserve threatened amphibians, the obvious target should be 100% of species. I suggest revising the manuscript accordingly.

2) Data. I don't understand why the authors use ecoregions as their base unit for spatial analysis, when data at much finer resolution are available as species by species range maps (http://www.natureserve.or...). Using these mapped across, say, equal-area grid cells would deliver much more informative results. In addition, I don't understand why the authors use data only 697 threatened amphibians, rather than 918 (the number derived from a search for threatened amphibians in Mesoamerica and/or South America, on www.iucnredlist.org).

3) Analytical techniques. As the authors point out, their use of heuristic area-selection techniques likely does not result in serious area-inefficiencies in terms of number of regions selected compared to optimal values. However, a real problem of suboptimality does occur, because the authors consider the specific identity of selected areas, and the distribution of these relative to other selected areas (e.g., for AL vs TD). This neglects the fact that any given set of areas selected by a heuristic solution represent just one of many - maybe very many - equally efficient sets of areas. This problem could be overcome by calculating and comparing irreplaceability (see extensive literature by Pressey and others) rather than minimum sets.

In addition, the discussion is rather rambling, and could be shortened considerably; and a careful spell check is needed throughout.