Abstract.− Biodiversity conservation design, though difficult with fragmentary or insufficient biological data, can be planned and evaluated with several methods. One of them, the complementarity criterion, is commonly used to account for the distributions of a number of species (i.e., an autoecological approach). At the same time, the patchiness and spatial bias of available distribution data has also been dealt with through distribution modelling. However, both the uncertainty of the ranges estimated and the changes in species' distributions in response to changing climates, limit the potential of single-species distributions as the biodiversity attribute to be used in complementarity strategies. Several technical and theoretical advantages of composite biodiversity variables (i.e., a synecological approach) may, however, make them ideal biodiversity indicators for conservation area selection. The drawbacks associated with current biodiversity data are discussed herein, along with the possible advantages and disadvantages of conservation planning through a synecological or autoecological approach.