Proximity, Interpenetration, and Sympatry Networks: A Reply to Dos Santos et al.

M. DOLORES CASAGRANDA; J. SALVADOR ARIAS; PABLO A. GOLOBOFF; CLAUDIA A. SZUMIK; LEILA M. TAHER; TANIA ESCALANTE; JUAN J. MORRONE

SYSTEMATIC BIOLOGY

Oxford Journals

Lugar: New York; Año: 2009 vol. 58 p. 271 - 276

1063-5157

In a recent paper, Dos Santos et al. (2008) criticize current methods to identify areas of endemism, including those described and implemented by others (Morrone 1994; Szumik et al. 2002; Goloboff 2004; Szumik and Goloboff 2004, 2007), and propose a newmethod. According to Dos Santos et al.,"Areas of endemism are traditional units for historical biogeography. Several approacheshave been proposed to identify and delimit these basic units of analysis in recent years,but all have particular methodological problems." (Dos Santos et al. 2008, p. 432)To overcome those ”methodological problems,” they propose a new method that uses as data the individual records because ”in the process of obtaining the ranges, either with grids or vectorial alternatives, the information provided by direct evidence (punctual records) ismasked” (Dos Santos et al. 2008, p. 446). They claim (Dos Santos et al. 2008, p. 447) that their method ”yields results strictly adjusted to a notion of endemism in thesense of species restricted to an area.” Whereas the criticisms to some of the methods theycite may be valid, others are hardly relevant or constitute only misrepresentations. In addition, their new techniques are—as we discuss below—based on substantialalterations to the traditional notion of endemism and use some of the very assumptions they criticize in other methods. As a consequence, the areas identified by their new method often correspond to nothing that could be sensibly called an ”area of endemism.”This is illustrated with several hypothetical examples analyzed with the implementation provided by Dos Santos, the R package SyNet (available at http://www.cran.r-project.org).