Little is known about the heritable behavioural traits of attacks by large carnivores on people. During the last 30 years attacks by brown bears Ursus arctos on people in the Cantabrian Mountains of Spain have been disproportionately concentrated in the eastern subpopulation. Excluding factors such as the existence of a single unusually bold bear, a higher human population density, particular human activities promoting encounters, or clear habitat differences in the area of this subpopulation, we propose that a plausible explanation for the unbalanced geographical attack pattern is that this subpopulation, separated a century earlier from the western subpopulation, may harbour a higher proportion of bolder bears. In the absence of genetic analyses this explanation remains speculative, but supports the hypothesis that genetic variation on the shy–bold continuum may influence attacks of large carnivores on people.

Reintroductions are challenging, and success rates are low despite extensive planning and considerable investment of resources. Improving predictive models for reintroduction planning is critical for achieving successful outcomes. The IUCN Guidelines for Reintroductions and Other Conservation Translocations recommend that habitat suitability assessments account for abiotic and biotic factors specific to the species to be reintroduced and, where needed, include habitat quality variables. However, habitat assessments are often based on remotely-sensed or existing geographical data that do not always reliably represent habitat quality variables. We tested the contribution of ground-based habitat quality metrics to habitat suitability models using a case study of the swift fox Vulpes velox, a mesocarnivore species for which a reintroduction is planned. Field surveys for habitat quality included collection of data on the main threat to the swift fox (the coyote Canis latrans), and for swift fox prey species. Our findings demonstrated that the inclusion of habitat quality variables derived from field surveys yielded better fitted models and a 16% increase in estimates of suitable habitat. Models including field survey data and models based only on interpolated geographical and remotely-sensed data had little overlap (38%), demonstrating the significant impact that different models can have in determining appropriate locations for a reintroduction. We advocate that ground-based habitat metrics be included in habitat suitability assessments for reintroductions of mesocarnivores.

Cactaceae is one of the most threatened plant families, in part as a result of the illegal extraction of plants for ornamental use. However, reports of the seizure and reintroduction of cacti are scarce and do not include species of Melocactus, the genus of Cactaceae in Brazil that has the highest number of threatened species. The coroa-de-frade Melocactus violaceus is endemic to Brazil and categorized as Vulnerable on the IUCN Red List. We report the seizure of 37 individuals of coroa-de-frade extracted illegally from their natural habitat, the results of their reintroduction to Paulo César Vinha State Park, in Espírito Santo state, Brazil, and provide information for environmental monitoring agencies regarding how to proceed in seizure cases, with the goal of minimizing the impacts of this illegal practice on the species. After seizure, 25 individuals were cultivated in a greenhouse and 12 were reintroduced in restinga, the natural area of occurrence of the species. After 6 months, survival was 76% for those individuals cultivated in the greenhouse and 84% for those planted in restinga, showing that rapid reintroduction of species with ornamental appeal, preferentially in their natural habitat, can reduce the impacts of illegal extraction. This reintroduction protocol can be used by managers of conservation units, contributing to the maintenance of threatened cactus species in their natural habitat.