Avifauna montana do Pantepui: modelos de distribuição de espécies e estudo de caso na Serra da Mocidade, Roraima, Brasil

Abstract

Located on the Guiana Shield, the Pantepui is a biogeographic region formed by tabular mountains, known as tepuis, of great scientific interest due to their endemic biota and notorious inaccessibility. Ecologically, Pantepui incorporates other types of mountain ranges, with distinct shape and geological origin, but which harbor species considered typical of this region. The majority of Pantepui is located in Venezuela, but it also expands to bordering regions of Brazil and Guyana and isolated components in Colombia and Suriname. After almost two centuries of ornithological studies, knowledge about the Pantepui avifauna is still far from complete. Ornithological expeditions have historically prioritized large tepuis, usually in Venezuela; however, recent studies have revealed the presence of a typical tepui avifauna in smaller and more isolated mountain ranges, increasing knowledge about the distribution of these species. The main objective of this work is, based on points of occurrence extracted from an extensive bibliographical survey and cietific collections, to model the distribution of montane bird species to predict which localities, within the geographical extension of Pantepui, have high environmental suitability for the occurrence of these species. Subsequently, we tested the accuracy of the models by comparing the predicted values of environmental suitability with empirically obtained field records during an expedition to Serra da Mocidade, a previously unexplored mountain range in the state of Roraima, northern Brazilian Amazonia. Besides the occurrence of tepui species, the Serra da Mocidade presents interesting biogeographic characteristics such as the longitudinally central location in Pantepui and its pronounced isolation, surrounded by lowlands, favoring the occurrence of endemic species that, since it is a place never studied before, would necessarily represent taxa new to science. We used the MaxEnt algorithm to model the distribution of 94 species considered characteristic of Pantepui, those absent from the adjacent lowlands or endemics of one or more of the mountains. To generate environmental suitability, we used a package of bioclimatic variables and altitude. In order to compare the predicted results with the records obtained in Serra da Mocidade, we extracted from the models the maximum suitabilityix value in the area sampled during the expedition. We defined a threshold of suitability to determine the species expected to occur in the Serra da Mocidade. Of the 121 typical Pantepui bird species, it was possible to model 94, of which 41 were recorded on Mocidade; all but three of these were predicted to occur. Of the other 53 Pantepui species not found on Serra da Mocidade, many were predicted to occur. We believe that most of these absences are related to the lack of appropriate habitat on Serra da Mocidade, and therefore, to the lack of adequate variables in the model, or to the intrinsic dispersion capacity of the species, which becomes unable to colonize Serra da Mocidade due to its high degree of isolation. This study adds knowledge about the montane bird species of Pantepui, a region that is very little studied relative to the lowlands that surround it, mainly due to the great difficulty of access. The results highlight the importance of empirical validation on spatial projections of species distribution models. It also emphasizes the relevance of an island biogeography perspective and the importance of studies in smaller or more isolated mountains of this biogeographic region, since these mountain ranges affords important tests of dispersal and colonization capacity, provide information on habitat preferences, and potentially enlarge the distribution of species, thus increasing knowledge of the natural history of the entire Pantepui.