Predicting the tangible and intangible costs of co-occurring with wildlife

摘要

Perceived intangible costs of co-occurring with wildlife are one of the strongest drivers of attitudes towards wild species, yet they have been little studied. Thus, determining the factors that influence the perception of such costs is essential to devise conservation strategies in landscapes where human-wildlife interactions are high. The main aim of this paper was to compare the effect of socio-demographic, exposure and attitudinal factors on the perceived cost probability of two species with associated tangible and intangible costs (i.e. crop-damage and fear respectively). Data were collected through semi-structured interviews and analyzed through logistic regressions under Bayesian approach. Model assessment and comparison was conducted using leave-one-out cross-validation (LOO) and their expected log pointwise predictive density (elpd). Although for both species, the exposure model was favored over the socio-demographic and attitudinal models, their perceived cost probability exhibited opposite responses. In the case of the crop-damaging species, the exposure model had considerably better predictive performance than the other two models. However, in the case of the feared species, the predictive performance of the exposure model was not better than the predictive performance of the other models. These results suggest: first, that the drivers of the perceived cost probability of co-occurring with wildlife are cost-specific; and second, that the perceived cost probability of the feared species has a lower predictability than the perceived cost probability of the crop-damaging species, given the set of candidate models. Considering that fear is an important perceived intangible cost of co-occurring with wildlife, more research is needed on its drivers, as well as on strategies to minimize it.