LAUSR

With colleagues, I critiqued reductive, polygon-based methods for estimating species’ geographic ranges (Peterson et al. 2016). This approach was used in numerous on-the-ground conservation-planning exercises (see, e.g., Ocampo-Peñuela et al. [2016] and references therein), and I appreciate the authors’ desire for practicality and agree that identification of potential habitat based on remotely sensed data may improve local-scale planning and decision making. However, I am concerned about the robustness of their mapping approaches in finescale applications to conservation planning. My colleagues and I presented a case study (Blackthroated Jay [Cyanolyca pumilo]) that we explicitly described as “a partial, preliminary example” (Peterson et al. 2016). Pimm et al. (2017) correctly point out that our occurrence data could have been vetted more carefully and model predictions explored in greater detail. A fully developed model, however, was not our goal. Rather, we wished to present an example of a data-driven approach as a counterpoint to their assumption-driven approaches. Unfortunately, in their rebuttal, Pimm et al. (2017) focused entirely on the example rather than on our general points. I therefore comment on 3 general points: the appropriateness of their use of IUCN extentof-occurrence maps as a starting point; problems with the assumptions their method requires; and the perils of developing such analyses at overly fine spatial resolutions.