El uso de modelos predictivos de distribución para el dise？o de muestreos de especies poco conocidas

the usefulness of predictive distribution models to provide a reliable representation is seriously questioned in the case of species with few well known localities. however, we propose in this paper that these techniques may be valuable for the design of future samplings devoted to collect new observations of little known species and, with limitations, to evaluate the possible extent of decrease in the original distribution area due to land use changes. with three different modelling techniques that use only reliable presence data, this study identified favourable areas with the capacity to support undetected populations of the lesser known species of no-flying mammals in the province of misiones, argentina. the probable degree of contraction in potential distribution area due to land use changes is also estimated.