Modelagem de distribuição de espécies de plantas raras e ameaçadas na savana amapaense

Abstract

The Amapá savanna is the second largest continuous savanas area in the Amazon and faces pressures from agricultural expansion, forestry, and fires, in addition to low representation in Conservation Units. This scenario increases the risk for rare and under-sampled species and reinforces the need to better understand their distribution. This study aimed to estimate the potential distribution of four species associated with the Amapá savanna, identify relevant environmental constraints, and verify the coverage of these areas by protected areas. The models integrated occurrence records from public databases and bioclimatic variables from WorldClim 2.1 in maximum entropy models in R, calibrated on the state's savannah region. The evaluation considered a 5 percent omission threshold, the binomial omission test, and percentiles of presence, allowing comparability even with reduced sample sizes. The models showed patterns consistent with the species' ecology and indicated that significant portions of the most suitable species remain outside of Conservation Units, highlighting protection gaps. The significance analysis identified thermal and water seasonality as determinants of environmental suitability. The most suitable nuclei often occur in pasture matrices, forestry and agricultural mosaics, close to highways and settlements, which suggests greater exposure to vegetation suppression and fire. It is concluded that the maps produced provide a technical basis for guiding field validation campaigns, supporting the creation or expansion of protected areas, and informing formal risk assessments for the savanna flora of Amapá.