Abstract
The concept of the ecological trap, a low‐quality habitat that animals prefer over other available habitats of higher quality, has appeared in the ecological literature irregularly for over 30 years, but the topic has received relatively little attention, and evidence for traps remains largely anecdotal. Recently, however, the ecological trap concept has been the subject of a flurry of theoretical activity that is likely to raise its profile substantially, particularly in conservation biology. Ecological trap theory suggests that, under most circumstances, the presence of a trap in a landscape will drive a local population to extinction. A number of empirical studies, almost all of birds, suggest the existence of traps and demonstrate the difficulties of recognizing them in the field. Evidence for ecological traps has primarily been found in habitats modified by human activities, either directly (e.g., through the mowing of grassland birds' nests) or indirectly (e.g., via human‐mediated invasion of exotic species), but some studies suggest that traps may occur even in relatively pristine areas. Taken together, these theoretical and empirical results suggest that traps may be relatively common in rapidly changing landscapes. It is therefore important for conservation biologists to be able to identify traps and differentiate them from sinks. Commonly employed approaches for population modeling, which tend to assume a source‐sink framework and do not consider habitat selection explicitly, may introduce faulty assumptions that mask the effects of ecological traps and lead to overly optimistic predictions about population persistence. Given the potentially dire consequences of ecological traps and the accumulating evidence for their existence, greater attention from the community of conservation biologists is warranted. In particular, it is important for conservation biologists and managers to incorporate into conservation planning an explicit understanding of the relationship between habitat selection and habitat quality.