There are roughly 300 species of poison frogs in the superfamily Dendrobatoidea (Dendrobatidae + Aromobatidae), and the rate at which new species are being discovered shows no signs of slowing down. Grant et al. (2006) published a major revision to the taxonomy of poison frogs. While we accepted many of the changes he made (especially in the family Aromobatidae), we also had (initially) many concerns regarding the revision of the family Dendrobatidae, particularly of the subfamily Dendrobatinae. Our primary concern was that the phylogenetic methods used in Grant et al. (2006) are somewhat controversial, and were therefore concerned that future phylogenetic studies might yield different results. However, we have addressed this topic in recent studies, and found that the topology presented by Grant et al. (2006) is robust to various methodological differences. Therefore, the monophyletic clades identified by Grant et al.’s study are stable, and thus the phylogeny upon which the taxonomy rests is now largely uncontroversial.

The current controversy regarding poison frog names is mainly subjective. Should poison frogs be retained as one large family, or split into two? Should Dendrobates be considered a single genus containing 37 species or split into 7 genera containing anywhere between 1–16 species? Higher-level taxa like genus and family have no accepted underlying “concepts” (unlike species): the only real requirement is that they be monophyletic. In general, taxonomic “splitting” increases precision of names and can facilitate communication. For example, it is easier to say Oophaga than “obligate egg feeders within the genus Dendrobates”. However, splitting also has the drawback of causing taxonomic instability. For example, someone reading an older paper or performing a literature search may fail to realize that Dendrobates arboreus and Oophaga arborea are the same organism. Thus, the answer to the questions posed above is largely a matter of taste, and of utility. If a newly proposed taxonomy is scientifically sound, and it is useful, then it should be accepted by the scientific community. We have found the current taxonomy to be useful, and thus have adopted it throughout this website and our publications.

Below is an outline of the taxonomy proposed by Grant et al. 2006. Major changes include (1) the family Dendrobatidae has been split into two families: Aromobatidae and Dendrobatidae, (2) the former Colostethus has been restricted to only 19 species, and the genera Allobates and Hyloxalus contain nearly half of all dendrobatoids, (3) the genus Ameerega refers to essentially the old Epipedobates minus the tricolor group, (4) Dendrobates has been restricted to only 5 species (primarily the large non-egg feeders), (5) Oophagarefers to the histrionica group (obligate egg feeders), and (6) Ranitomeya refers to the tiny Amazonian “thumbnail” species.

More recent updates include the recognition of Excidobates (Twomey & Brown 2008) and Andinobates (Brown & Twomey et al. 2011).

Clicking on genus names will bring up a short description of that genus, a list of species included in the genus, and links to field accounts and original publications (if available). For now, we are only focusing on frogs of the family Dendrobatidae. More information on Aromobatidae can be found on the Amphibian Species of the World website.