Influence of data sampling and structure on areas of endemism

PABLO A. GOLOBOFF; CLAUDIA A. SZUMIK; CASAGRANDA M. DOLORES

St. Petersburg, Florida

Congreso; XXXVI Hennig Meeting; 2017

Willi Hennig Society

Properties of distributional datasets -- such as incomplete sampling, sample bias,and imperfect detection-- are widely recognized to be a key factor onbiogeographical inferences and identification of biogeographic patterns. The impact of these has been explored in ecological models, but consistently ignored in historical biogeography. Understanding the influence of these factors could give us important information to evaluate the consistency or support of empirical results. Here, we simulate different kinds of bias on real datasets in order to analyze their effects on the identification of areas of endemism. We use different types of resampling, simulating alternative types of bias (or incomplete sampling) commonly present in distributional databases. We compare the results of resampled data with those obtained with the complete data and calculate patterns of dis/similarity. This study shows the incidence of data structure on recognition of areas of endemism, shedding light on this problem and taking us closer to support measures for hypotheses of areas of endemism.