October 22nd, 2006 - Dendrobatid systematics

In August 2006 Grant et al. published a large manuscript concerning the systematics of the family Dendrobatidae. This manuscript contributes a huge amount of genetic and morphological information to dendrobatid systematics. However, we are choosing to be conservative and use the traditional nomenclature for Epipedobates and Dendrobates. We will however be honoring all other proposed genera as they dramatically improve the current taxonomy.

Below are some of the reasons why Dendrobates.org has not yet accepted the proposed taxonomy of Dendrobates and Epipedobates:

1. The splitting of Dendrobates into Adelphobates, Oophaga, Ranitomeya, and Minyobates is not usable in the field. A working taxonomy should use characters which are distinguishable between groups other than simply where the organism was found. The proposed genus Adelphobates has no novel distinguishing characteristics from Oophaga, Ranitomeya, or Dendrobates. Further, all proposed characters in absence of molecular phylogenetics place all genera into one mixed strongly monophyletic group. Although molecular phylogenetics can consistently define groups that form these proposed genera, the relationships between groups is not being consistently recovered (compare Grant et al. 2006, Vences et al., Noonan  and Wray 2006, Roberts et al. in press) and therefore the relationships of these groups a far from resolved.

2. Names proposed by Bauer (Oophaga, Ranitomeya, Ameerega) only partially meet ICZN requirements; only in his 1986, 1988, and 1994 publications do the names barely meet the minimal requirements (basically naming a type species, which has little to do with utility of the proposed genus). Publications concerning taxonomy, especially genus-level taxonomy, should be performed by taxonomists. When this work is published it must be peer reviewed and published in the appropriate scientific journals. None of Bauer’s publications which propose these changes were peer reviewed and all were privately published. Lastly, a note considering gender changes: although not an explicit requirement, it does facilitate and simplify taxonomy if new genera remain the same gender as the previous genus. For these reasons we question the validity of Bauer’s proposed genera, and have decided to not honor these names until this issue has been investigated further and the scientific community comes to an agreement.

3. We realize that the Epipedobates is in fact paraphyletic and contains (at least) two genera: the tricolor group and “all other Epipedobates”. Although Grant’s taxonomy does fix this problem by restricting Epipedobates to the tricolor group and using Ameerega to refer to all other Epipedobates, the validity of the later name is an issue of contention, as described in point 2.  Other valid, although not perfect, and peer reviewed publications concerning the naming of the genus of “all other Epipedobates” have been published around the same time as Bauer’s 1986 publication. Zimmerman and Zimmerman (1988) proposed the name Phobobates, which we feel ameliorates many of the problems presented in 2. Of course, we do realize a genus name has rules of precedence and can not be chosen simply to suit convenience.

4. Purely parsimony based phylogenetic inferences. Sometimes the simplest scenario isn’t the best. This Occam’s razor-based method isn’t the most widely used or most robust method for inferring phylogenies (see http://en.wikipedia.org/wiki/Maximum_parsimony), and Grant et al. neglected to use other methods. A thorough phylogenetic analysis would use all three main methods (Parsimony, Maximum Likelihood, and Bayesian inference) and build a taxonomic hypothesis based on the agreement of all methods, or at least justify why one method is preferred over the other, while presenting results from the other methods. A pious devotion to any one phylogenetic method is unscientific, and dangerous when used to make broad taxonomic changes.