Landscape of amphibian diversity in Latium Region: peaks, valleys and gaps of conservation priority

Abstract

Faunistic inventories of particular regions would provide essential datasets for conservation and resource management, but often data are not evenly distributed through the space and time as a consequence of an unequal sampling effort. Therefore, to accurately estimate species diversity it is necessary to minimize this potential confounding effect embedded within most existing datasets. In this study, we identified as high-priority conservation targets areas of high-value species diversity of amphibians in Latium (Central Italy), by illustrating a methodology to generate biodiversity patterns from very heterogeneous datasets, which accounts for the relationships between species richness and sampling effort. We analyzed 6656 data records collected from 1991–2011, which were assigned to the corresponding 10 × 10 km cell of the UTM grid. In order to remove the bias introduced by the differential field effort across cells, we used regression analyses after applying a curve estimation procedure. Then, we validated the best regression model by comparing species richness estimated using this model against that obtained using rarefaction curves. Our results show that the use of residuals can be a suitable approach for controlling the influence of sampling effort on the observed species richness except in a very unequal distribution of the records. Indeed, residuals can be calculated for any cells, irrespective of the observation effort, while when using rarefaction curves a relatively high number of records per cell is required. Finally, when contrasting the amphibian diversity pattern with the existing network of protected areas, several conservation gaps are identified in Latium, especially in mountainous areas.