Finding a lost species in the ‘Lost World’: predicted habitat occupancy by an endemic butterfly in a Neotropical sky‐island archipelago

Abstract

Abstract Pantepui is a Neotropical archipelago of remote sky islands (tepuis) that harbours a unique and poorly known biota, such as the endemic butterfly Antirrhea ulei. The Vicariance‐Migration hypothesis argues that Pantepui biota originated from a complex succession of climatic shifts, causing up‐and‐down migrations of cool‐adapted species from the last glacial maximum (LGM) to the present. We evaluated how environmental gradients affect A. ulei habitat occupancy and predict its distribution across the Eastern Pantepui. We also test whether the species had a broader distribution during the LGM, following the Vicariance‐Migration hypothesis. We surveyed for butterflies across 14 plots at Uei tepui, following an elevational gradient. We used Bayesian occupancy modelling to evaluate how environmental gradients affect A. ulei occurrence and to predict the current and past species distribution. Species habitat occupancy was strongly associated with environmental gradients. Our model correctly predicted the species occurrence at all localities previously reported to be occupied by the species, and also predicted the potential species occurrence on other tepuis. Our historical prediction of species distribution showed that the species likely had a broader distribution during the LGM, in comparison with its current restricted distribution. Our historical predictions suggest that the species may have spread across the Eastern Pantepui during LGM and migrated up the tepuis during the Holocene warming, in accordance with the Vicariance‐Migration hypothesis. Our study shows how data from local standardised surveys can be useful to estimate the distribution pattern of other little‐studied species of the Pantepui biota.