Fifty years after: A taxonomic revision of the amphibian species from the Ecuadorian biodiversity hotspot Abra de Zamora, with description of two new Pristimantis species

that P. colodactylus may be related to P. proserpens and P. celator, and in 1980 Lynch and Duellman [59] assigned it to the P. celator assemblage of the P. unistrigatus species group. The membership to the P. unistrigatus species group was maintained by Duellman and Lehr [25] and Hedges et al. [41]. With the description of P. muranunka, Brito et al. [29] recognized the morphological similarities of this species with P. colodactylus (and also with P. proserpens, P. paquishae, and P. tinajillas). In 2019, Urgilles et al. [16] recognized P. colodactylus (together with P. vidua and P. tinajillas) as a potential member of the P. orestes group. We provide genetic evidence that P. colodactylus is a member of the P. orestes group, which, together with P. muranunka, P. matildae and an undescribed species, forms a subgroup of closely related, morphologically distinct, bromeliad specialist species (Fig 7). In the original description of the species Lynch [9] observed several similarities between P. balionotus and P. percultus and “very tentatively” associated them with the balionotus-riveti- ruidus series. In 1980, Lynch and Duellman [59] assigned both species to the P. unistrigatus species group, although to two different assemblages: P. balionotus to the P. unistrigatus assem- bly and P. percultus to the P. glandulosus assembly. Hedges et al. [41] assigned both species to the P. unistrigatus species group. In 2019, Pa´ez and Ron [30] assigned P. balionotus and P. per- cultus, based on their similar morphology, to the P. phoxocephalus species group of the newly

proposed Huicundomantis subgenus. Our analysis support their assumption and show that both species are members of the P. phoxocephalus species group. In our phylogram, P. baliono- tus kinship to the other members of the group remains non-resolved (with suggested relation- ship to P. hampatusami and an undescribed species), but P. percultus is part of a strongly supported subgroup composed of P. chomskyi, P. andinogigas, P. atratus and P. multicolor (Fig 21). In 2019 Ya´nez-Muñoz et al. [60] described a new, “giant”, Pristimantis species, P. andinogi- gas, from the Cajanuma sector of Parque Nacional Podocarpus. As their description was not based on a phylogenetic analysis, the authors refrained from assigning it to any species-group although they noted its morphological similarity to members of the former P. orcesi group. In our analysis, P. andinogigas is closely related to P. chomskyi and to other three species (P. per- cultus, P. atratus and P. multicolor) of a strongly supported subgroup of the P. phoxocephalus species group. However, a careful revision is needed for P. andinogigas and P. chomskyi because the genetic distance between the two species is around 2% and it seems that the description of P. chomskyi was based on subadult specimens, as the individuals from the type locality and nearby areas are in fact large, morphologically very similar animals to P. andinogi- gas (pers. obs.). Conservation remarks Even though most of Abra de Zamora is well preserved and about 50% of its total area is pro-

tected, being inside Parque Nacional Podocarpus, the amphibians of this area face several threats. Among the main threats we can mention are the loss and degradation of habitats, due to cattle farming, and the introduction of exotic species (Rainbow Trout, Oncorhynchus mykiss and Bullfrog, Lithobates catesbeianus), species that have already been proven to have a direct negative effect on Andean amphibian populations [61]. On the other hand, forest fires are also a potential threat, since they occur frequently in nearby sectors intended for agriculture or livestock. As for the infectious diseases that are affecting the Ecuadorian amphibian populations, we have limited information from the study area. The presence of chytrid fungus (Batrachochy- trium dendrobatidis) is documented from the Andes of central and northern Ecuador [62, 63]. A recent survey conducted by researchers from University of Central Florida, yielded negative PLOS ONE Taxonomic revision of the amphibian species from Abra de Zamora PLOS ONE | https://doi.org/10.1371/journal.pone.0238306 September 10, 2020 54 / 59