Targeting the survey efforts: Gaps and biases in epiphyte sampling at a biodiversity hotspot

Species distribution data are essential to understanding and protecting biodiversity against current human threats. However, biodiversity knowledge is unevenly distributed geographically and needs to be evaluated to target deficient knowledge locations and support conservation planning. In this study, we intend to assess the inventory completeness of the vascular epiphytes from the Atlantic Forest, in addition to: comparing the sampling effort between (i) phytophysiognomies, (ii) epiphyte taxa, and quantifying (iii) spatial and (iv) environmental biases. We used the average of the last 10% of Species Accumulation Curves (SAC) as the inventory completeness and quantified their spatial biases using Spatial Autoregressive (SAR) models. We evaluated the environmental biases using the Mann-Whitney U test. We found that most epiphyte inventories in the Atlantic Forest are undersampled, with well-sampled inventories concentrated on the southeastern coast; thus, Ombrophilous forests are better sampled than other phytophysiognomies. These well-sampled inventories are spatially biased mainly in large forest cover, protected and low agricultural coverage areas. In addition, epiphytes were more collected in regions environmentally associated with high diversity, such as areas with low temperature, high rainfall and altitudes. More conspicuous and richer groups had the highest sampling coverage, but the lowest effort when compared to poor groups. In order to reduce the gaps, we suggest that future inventories focus on the continental portion of the Atlantic Forest, since these areas are already highly threatened by land-use and have large Wallacean shortfalls for all epiphyte taxa. Therefore, enabling future researchers to propose efficient conservation actions in these priority areas.