In addition to the usual suite of human-related threats to persistence (habitat destruction and fragmentation, competition and exotic diseases from livestock, poaching), the conservation of endangered wild sheep and goats (Ovis and Capra spp.) is hampered by a limited knowledge of their taxonomy. Taxonomic knowledge is important because it allows the preservation of local adaptations, avoidance of hybridization and, for reintroductions, the re-establishment of the right species in the right places. The taxonomy of wild sheep and goats, especially in Africa and Asia, is complicated by intergradation and hybridization of taxa, limited data owing to poor accessibility to much of their remote mountainous habitat, and contrasting results of molecular and morphological studies. Of all wild goats, the least known and the most endangered is the Walia ibex Capra walie, restricted since historic times to the Simen Mountains of Ethiopia, where a current population of about 450 uses 95 km of rugged high-elevation habitat. Although these ibex look very different from Nubian ibex Capra nubiana found in deserts around the Red Sea, they are likely related to them, and their status as a distinct species has been questioned. Gebremedhin et al. (2009) used DNA extracted from fecal samples and information on the physical and climatic characteristics of habitat used by both Walia and Nubian ibex to conclude that there are considerable genetic and substantial ecological differences between these taxa. Those differences justify the treatment of Walia ibex as a separate and valuable conservation and taxonomic unit. In other words, Walia ibex are a ‘good species.’ Unlike the desertadapted Nubian ibex, the Walia ibex lives in a cold, highelevation mountainous environment with substantial precipitation. Even though the two groups probably diverged less than a million year ago, they are now very different. As the authors point out, the estimated time since these two ibex became separated is likely three times as long as the apparent divergence time between polar and brown bears (Ursus maritimus and Ursus arctos). Clearly, morphological differences and ecological adaptations are not necessarily a simple function of time or of genetic divergence: this is an extremely important point for both ecology and conservation. The main key to taxonomic divergence and therefore to speciation is not time, but differences in local selective pressures. Those different selective pressures lead to different adaptations, and preserving local adaptations must be a major goal of conservation. That means that we cannot simply quantify genetic differences along some chunk of DNA and always expect those differences to closely match either ecological specialization or conservation value. Both taxonomy and conservation require a holistic approach, considering genetics, morphology, ecology and, above all, evidence of local adaptation. The key point here is not whether Walia ibex either looks different from Nubian ibex or whether the Fst between the two is large or small – the point is that Walia ibex evolved adaptations to the highlands of Ethiopia and Nubian ibex are adapted to the desert! Pretending that weak genetic differentiation is relevant to conservation in this case is like arguing that polar and brown bears are not different. Importantly, the research by Gebremedhin and colleagues also identifies potential suitable habitat for Walia ibex outside its present range. These areas are possible candidates for introduction or reintroduction. Unfortunately, there is currently no information available about the former distribution of Walia ibex outside the Simen Mountains, and the study also concluded that this species’ decline may have been ongoing for a very long time. Knowledge about potential habitat, however, is very valuable for the conservation of this species, as the establishment of a separate population may serve as an insurance policy against any catastrophe that may affect the Simen population. On the other hand, habitat analyses cannot by themselves identify all threats to a population, and before any introduction or reintroduction is attempted, information is required about competition from livestock, risks of poaching and other possible threats that may be faced by a new population. Those considerations are particularly important for a species that is already at risk because of low numbers, very small distribution and, as Gebremedhin and colleagues report, low genetic variability. Inevitably, any newly established population would have less genetic variability lower than the source. However, because the study found that