Comparison of resolution of methods used in mapping biodiversity patterns from point-occurrence data

This paper examines three methods of mapping of biodiversity using point-occurrence data for the birds of Mexico: aggregation of species occurrence records, vegetation surrogate, and individual species models. We compare the approaches from the perspective of achieving potential gains in spatial resolution with existing data. We found that mapping the diversity of Mexican birds using individual species models yielded results 400-fold more finely resolved, quantifiable errors, and greater flexibility for many applications. We show that the aggregation and surrogate methods are susceptible to tradeoffs between bias and resolution that can only be ameliorated thorough more intensive sampling. A theoretical error model and an empirical demonstration shows that higher spatial resolution in the individual species approach can be achieved by controlling the modeling approach by reducing bias and decreasing random error. The method is particularly applicable for large-scale biodiversity mapping, where intensive ground survey data are lacking.