Effect of chronological addition of records to species distribution maps: The case of Tonatia saurophila maresi ( C hiroptera, P hyllostomidae) in S outh A merica

Ecological niche models have become very popular for analysing the potential distribution of species. Nevertheless, models are strongly influenced by many factors, such as spatial resolution, environmental variables and the quality of distribution records. In this paper, we evaluated how ecological niche models changed with the addition of records accumulated over four decades. Our model species was the stripe‐headed round‐eared bat ( T onatia saurophila). Thus, with data organized in chronological order, we could observe how the models changed in predicting distributions over time in comparison with all known point locations. We tested if partial models could predict the occurrence of new unpublished records for savannah areas in central B razil, considering that the species is typically associated with forest environments. Our results indicate a high omission rate for models built with point localities from the 1970s and 1980s (58.5% and 50.0% of all known points respectively), and predicted that the species could occur in central B razil. Although T . saurophila has indeed been recorded recently in central B razil, it was found in places different from those predicted by the models using these restricted earlier data. Nevertheless, the environmental suitability of such areas is significantly different from sites largely described in earlier records from the A mazonia region, as shown by principal components analysis. We argue that populations of T . saurophila that occupy open habitats in central S outh A merica (including C aatinga, C errado, Chaco and semi‐deciduous interior forests) deserve further study at the genetic level to determine if bats in these very different habitats are taxonomically distinct from A mazonian populations. Our results also suggest that models based on very limited datasets for species occurrence can lead conservationists or decision makers to wrong conclusions.