Mountainous areas of the world are extremely important, among other things, because of the services their ecosystems provide and because they often constitute areas of high concentration of endemic species (Peterson et al., 1993; Jetz et al., 2004; Körner and Ohsawa, 2004). Despite the fact that the inventory of species on earth is far from fi nished (May, 1990), recent and misguided policies by many countries are making the collection of new specimens an increasingly diffi cult enterprise (Grajal, 1999). Although it is imperative that efforts to increase existing collections are maintained and strengthened, and to augment the taxonomic capacity in the developing world, it is paradoxical that access to the billions of specimens (Chalmers, 1996) that have already been collected was, until recently, very limited, and therefore, access to primary biodiversity data was restricted to a handful of experts. For scientists in the developing countries, it was diffi cult and expensive to have access to data in the specimens, because they are scattered over many institutions, countries, and continents. Besides, most specimens were not computerized, and even if they were, the data remained locked in the computers of the museums or herbaria (Soberón et al., 1996). Although some of the data was partially published in catalogues, monographs, and

other specialized work that often included the locality of the specimens used to do a revision or new descriptions, such specialized literature also tended to remain in the libraries of the major institutions of the North and therefore was almost inaccessible to nonexperts or to experts in the poor countries.