Resumen
Species distribution models are widely applied for generating the appropriate data for designing networks of reserve systems. The aim of the present study is to compare the use of presence/absence and continuous suitability data, derived from species distribution models, in reserve selection, and to detect any systematic trends in the reserve networks produced on the basis of these two types of data. The MAXENT model was applied to predict orchids? potential distribution in east Macedonia (northeast Greece). One presence/absence and one with suitability values data set was made and used in the ZONATION decision support tool in order to prioritize the cells of the study area for inclusion in a reserve network. In the selection procedure, species weighting has been applied by using the species specialization index. Results show that the presence/absence data set favors the selection of cells with more extreme climatic conditions and more distant from the mean habitat of the study area. Furthermore, the continuous suitability data set provides higher suitability values for the specialist taxa in comparison with the presence/absence data set, while the opposite is happening for the generalist taxa. The present study reveals that the suitability data outperform the presence/absence data in reserve selection because: (a) they better represent the average environmental conditions of the study area in the selected networks; (b) they ensure higher suitability values for the specialist species, which are more prone to extinction in the future; and (c) they take full advantage of species weighting according to their habitat specificity.