How do distribution mapping methods perform in estimating beta diversity at macroecological scales? A case study with Neotropical anurans.

Abstract

Species distribution mapping methods have their advantages and limitations concerning their use on theoretical and/or applied macroecological approaches. However, it remains underexplored how the estimates of community ecology metrics vary across the distributions generated by different mapping methods. Here, we mapped the distribution patterns of the anuran beta diversity in the Atlantic Forest and Cerrado hotspots generated by three mapping methods: point-to-grid (PTG), extent-of-occurrence (EOO), and ecological niche modelling (ENM) maps, so we were able to compare the congruence of the local contribution to beta diversity index (LCBD) among them, as well as their turnover and nestedness components. PTGs generated the most divergent LCBD values probably due to the more resolved spatial scale in which species' presence are considered, so EEO and ENM generated similar beta diversity estimates for both hotspots. High LCBD values in the Cerrado were recorded in ecotone regions, whereas in the Atlantic Forest the highest beta diversity values were found along the Atlantic coast. The structure of beta diversity of PTG showed way too high values of importance for the turnover component compared to the EEO and ENM maps, which also recorded higher importance for the turnover than for the nestedness component.