**1. Introduction**

Amphibians are experiencing a global conservation crisis, with an estimated 41 to 50% of species su ffering population declines [1–3]. These declines are likely the result of interactions among several factors of primarily anthropic origin, including overexploitation, habitat destruction, pollution, and the effects of emerging epizootic pathogens [4]. The most alarming cause of global declines and extinctions of amphibians is chytridiomycosis, a disease caused by the chytrid fungus, *Batrachochytrium dendrobatidis* (*Bd*) [5,6].

Harlequin frogs of the genus *Atelopus*(Anura: Bufonidae) have arguably suffered the most dramatic population declines and extinctions of any diverse genus of amphibians. Because of their diurnal habits, bright coloration, and previously high local abundance, species of this genus were a prominent element of many Neotropical communities until about 35 years ago [7–9]. Since the late 1980's, the majority of species of *Atelopus* have not been seen in their historic localities [10]. These declines were likely driven by *Bd*, and *Atelopus* are known to be highly susceptible to chytridiomycosis [11,12]. Declines of *Atelopus* and other amphibian species have been considered to be most severe in populations at higher elevations (above 1000 m above sea level) [13], hypothetically because the lower temperatures present at higher altitudes are associated with higher levels of infectivity and fecundity in *Bd* [14].

The eight species of Central American *Atelopus* are no exception to these patterns of endangerment and decline. These forest species have a very similar natural history, in which reproduction is associated with fast-flowing streams where eggs are laid in the water and tadpoles develop ([15], pers. obs.). The only exception is the scarcely known *A. chirripoensis* from the Costa Rican *páramo* which breeds in small ponds [16]. Central American *Atelopus* were once abundant members of amphibian communities at certain localities, particularly during their breeding season ([15,16], pers. obs.). Currently, all eight are listed as Critically Endangered by the International Union for Conservation of Nature (IUCN) and the three species restricted to elevations above 1000 m (*A. chiriquiensis, A. chirripoensis*, and *A. senex* [17]) are considered possibly extinct [18]. The much reduced populations of the remaining five species have been prioritized for *ex situ* conservation through captive breeding programs in Panama and abroad [19,20]. *Atelopus zeteki* is endemic to central Panama and it has not been seen in the wild since 2009. Individuals were collected before the epizootic declines, however, and this species is maintained in a well-managed captive breeding program [21]. The four species still found in the wild, *A. varius* of Costa Rica and Panama, plus *A. certus*, *A. glyphus*, and *A. limosus* of eastern Panama, are being reared in captivity by the Panama Amphibian Rescue and Conservation Project [22,23].

The success of *ex situ* conservation programs depends on having a complete and robust taxonomy, because correct delimitation and identification of species (1) allows a more efficient use of resources by avoiding the unnecessary protection of taxonomically invalid species [24], (2) can prevent neglecting previously unrecognized species in conservation programs [25], and (3) helps to avoid the *ex situ* generation of hybrids possibly maladapted to the environments inhabited by their parental species [26]. However, the taxonomy of Central American *Atelopus* is still incompletely resolved. Most species have been evaluated only based on characteristics of adult external morphology, and coloration pattern has been given special importance, even though it has been found that in *Atelopus* this character is not always concordant with genetic differentiation [27,28]. Additionally, the phylogenetic relationships of the species of *Atelopus* in Central America are drastically understudied, because only three species have been included in any published molecular phylogenies of the genus [10,28,29].

Given that Central American *Atelopus* seem to have a South American origin [10], understanding their phylogenetic relationships and historical biogeography could shed light on the timing of the closure of the Isthmus of Panama and its effect on animal lineages moving between continents. The term Great American Biotic Interchange (GABI) is sometimes used to refer to the massive interchange of mammalian lineages starting around 2.7 million years ago (Ma) [30]. Here, however, we use the term more broadly to refer to any exchange of fauna or flora between North and South America during the Neogene and Quaternary periods (cf. [31]). How the GABI relates temporally to the final geological completion of the Isthmus of Panama is a controversial topic, because the traditional date of roughly 3 Ma [32,33] has been challenged recently by Montes et al. [34] on geological evidence, and by Bacon et al. [35] on DNA sequence data, who propose a much older closing of the Isthmus around 15–10 Ma. In turn, O'Dea et al. [36] charged both studies with biased data collection and erroneous interpretation, and the controversy continues. Nonetheless, DNA sequence studies on amphibians and reptiles are often supportive of a closure date older than 3 Ma (e.g., [37–39]). In this work, the phylogenetic relationships of the species of *Atelopus* of Central America are estimated from mitochondrial DNA sequence data in order to evaluate the current taxonomic status of populations and species, as well as to investigate the number and timing of colonization events between South America and Central America.

## **2. Materials and Methods**
